MS-721 Collaboration Communications Systems Engineer Study Guide¶

Independent AI-assisted resource — HUMAN REVIEW PENDING. The October 28, 2026 team refresh reconciles the complete guide and detailed objectives with current public exam, product and learning evidence. See the current group report and sources-and-objectives record. Source conflicts and unexecuted exercises remain explicit. This is not human review or live product validation. The official blueprint is authoritative.

Study guide at a glance

Start here

Report an issue with MS-721

Blueprint checked2026-10-28

Review stateSources + objectives checked — human review pending

Guide-only active reading1 hr 30 min–2 hr 15 min

Prerequisites: Practical Microsoft Teams and Microsoft 365 administration plus networking, telecommunications, SIP/PSTN, audiovisual and meeting-room fundamentals; hands-on experience with Teams Phone, Direct Routing, emergency calling, meetings/events, Teams Rooms and devices, PowerShell, Call Analytics and CQD is strongly recommended.

Exam essentialsRead the objective map, key distinctions, and readiness checklist.

Deep understandingWork through every domain, including decisions, failure modes, and related items.

Hands-on labsBegin with Lab 1 — Architecture and licensing matrix, then use the remaining labs to produce evidence you can inspect and explain.

Official objective domains

Plan and design collaboration communications systems20–25%
Configure and manage Teams meetings, webinars, and town halls15–20%
Implement and configure Teams Phone30–35%
Configure and manage Teams Rooms and devices20–25%

The active-reading range uses 90–130 words per minute for technical material. Labs, troubleshooting, note-taking, spaced review, and prerequisite gaps add time; it is not a total preparation promise.

Retained teaching baseline: Skills measured as of April 28, 2026. The current live page displays the future October 29 outline; this does not freshly republish April.
Future outline — VERIFY CURRENT (October 28, 2026): The live study guide displays October 29 scope, which is future today. The accepted April outline and all future detailed objectives are mapped separately. The future wording distinguishes Teams Premium and Microsoft 365 Copilot, names appointments with Teams, and clarifies group policy assignment. Accepted objective/status bytes remain unchanged until an effective-date review; no early promotion is implied.
Official source: MS-721 study guide

The deep-review report records the evidence and limits. The current credential lists 100 minutes; practice vendors’ question counts are not promises about the exam.

How to use this guide¶

MS-721 is a systems-engineering exam. A feature is not finished when a portal toggle is on: its identity, license, number, carrier/SBC, policy, network, endpoint, emergency behavior, monitoring and support boundary must all agree. For each objective, practice this chain:

translate business, regulatory, room and user requirements into a supported topology;
choose licenses, identities, numbers, PSTN connectivity, devices and policy objects;
configure the smallest appropriate scope through Teams admin center or PowerShell;
validate inbound, outbound, failover, emergency, meeting and device behavior;
collect evidence in Call Analytics, Call Quality Dashboard (CQD), device portals and logs;
isolate the failing administrative, media, carrier, SBC, network, identity or endpoint boundary;
restore service and prove the expected and denied paths.

About related items: A Related item: callout adds prerequisite, operational, architectural, or adjacent context that makes the current topic easier to understand. It is useful supporting knowledge, not a claim that the item appears verbatim in the published exam objectives.

Objective map¶

Published domain	Weight	Central question
Plan and design collaboration communications systems	20–25%	Which meeting, voice, room and network architecture meets the requirement with a supportable boundary?
Configure and manage Teams meetings, webinars, and town halls	15–20%	Which tenant setting, policy, template, event setting and license controls the intended experience?
Implement and configure Teams Phone	30–35%	Can users and voice applications make, receive, route and survive calls safely?
Configure and manage Teams Rooms and devices	20–25%	Are shared endpoints correctly identified, enrolled, secured, configured, updated and monitored?

1. Plan and design collaboration communications systems¶

Meeting and event architecture¶

Select the collaboration format from audience, interaction, registration, production and control requirements—not just maximum attendance:

Experience	Best fit	Design questions
Meeting	Many-to-many collaboration	Lobby, presenters, chat, apps, breakout rooms, recording/transcription and content control
Virtual appointment	Scheduled customer or client interaction	Booking workflow, browser join, reminders, queue/analytics and advanced appointment capabilities
Webinar	Structured event with registration	Public versus organization scope, registration fields, presenters, Q&A, branding and attendee reporting
Town hall	Produced one-to-many broadcast	Producers/presenters, attendee scale, eCDN, moderated interaction, recording and publishing

Tenant meeting settings define service-wide capabilities/defaults. Meeting policies apply to organizers or participants depending on the setting. Templates package supported meeting options; template policies determine who can use them. Sensitivity labels can enforce protected meeting settings under applicable licensing. Per-meeting options finish the instance. Determine the controlling layer before changing global policy.

Plan polls, Forms, apps, Q&A, breakout rooms, avatars, controlled content, recording, transcription and Microsoft 365 Copilot together. Copilot behavior can depend on license, effective meeting policy, organizer options, transcript/recording choice, client and feature rollout. Teams Premium adds capabilities such as advanced meeting protection, templates, branding and enhanced event/appointment experiences; verify exact entitlement before promising a feature. The current Copilot matrix distinguishes the organizer-editable Disabled default from the enforced EnabledWithTranscript option. During-only speech processing does not imply that compliance copies of prompts/responses vanish. E2EE meetings do not support Copilot. Verify actual participant entitlements, options and transcript access in a rehearsal.

Audio Conferencing brings PSTN dial-in/out capabilities to meetings. Plan licenses, supported countries, default bridge number, toll versus toll-free inventory, dial-out restrictions and numbers shown in invitations. The bridge is a tenant service; a user-specific conferencing number and an organizer's policy/license still affect the invitation and behavior.

Related item: PSTN Audio Conferencing connects a caller to a meeting. Teams Phone connects a user or voice application to ordinary telephone calling. Both use numbers, but they are different services and license/configuration paths.

Teams Phone and PSTN connectivity¶

Set up Teams Phone separates the cloud phone service from the method used to reach the public switched telephone network. Select per user/site/persona where supported:

PSTN model	Carrier/infrastructure boundary	Strong fit	Main tradeoff
Microsoft Calling Plan	Microsoft supplies cloud PSTN service	Straightforward supported-region deployments	Availability, number and plan constraints vary by market
Operator Connect	Certified operator supplies PSTN and manages operator infrastructure	Retain a participating operator without customer SBCs	Operator geography/features and joint support boundary
Teams Phone Mobile	Participating mobile operator makes a SIM number the Teams number	Mobile-first users needing one business number	Supported operators/regions and user-number focus
Direct Routing	Organization/operator connects a certified SBC to Teams	Existing carriers, PBXs, analog devices, complex routing or multinational requirements	Organization owns more design, security, monitoring and support work

Calling Plans include domestic, international and pay-as-you-go choices where offered. Teams Phone Mobile is for user numbers; do not assume it supplies Audio Conferencing or voice-application numbers. A tenant can mix connectivity models, but every user/number needs an unambiguous operational and emergency-calling design.

Shared Calling lets eligible users make PSTN calls through a shared resource-account number rather than each receiving a dedicated number. Plan the resource account, number, emergency policy/location, voice routing and user calling policy. Shared Calling planning uses an auto attendant for inbound reachability and its number for outbound identity. Users still need Teams Phone and voice enablement. Configuration guidance requires the user’s ordinary voice-routing policy to have no PSTN usages, while a Direct Routing resource account needs a policy with the appropriate usages. An inherited global usage list can silently defeat the design. Optional extension dialing is a documented exception to the ordinary “no user number” setup. Moving the shared resource number leaves policy objects present but can break calls.

Inventory number purpose and ownership:

user/subscriber number: assigned to a voice-enabled user;
service number: used for services such as auto attendants, call queues or conferencing under the applicable model;
resource-account number: anchors a voice application or Shared Calling design;
private line: gives an eligible user another inbound route under current limitations;
SMS-capable Calling Plan number: supports current Teams SMS behavior only in supported markets/configurations.

SMS planning currently limits this service to eligible Microsoft Calling Plan users/numbers in the US (including Puerto Rico) and Canada, with approved Brand and Campaign registration even for individual business messages. Voice enablement alone is insufficient. Do not infer SMS support for Operator Connect, Direct Routing, toll-free numbers, shared queue inboxes or other clouds from ordinary calling support.

Worked example 1 — Count SMS segments, not chat bubbles. Suppose ten Domestic licenses contribute 200 segments each and two PAYG licenses contribute zero: the tenant pool is 2,000 segments. Measured outbound traffic contains 900 one-segment messages and 300 two-segment messages; inbound traffic adds 650 segments. Total usage is 2,150, leaving 150 segments over the pool. Use measured encoding/segment counts and current billing terms; the displayed number of messages or a simple character count can understate usage.

Document acquire versus port, losing/gaining carrier, authorization, emergency address, usage location, assignment date, rollback and validation. Porting is a business cutover: freeze conflicting changes, test temporary and final routing, and retain carrier escalation details.

Direct Routing, extensions and survivability¶

Direct Routing planning starts with a Microsoft-certified Session Border Controller (SBC), a public trusted certificate, supported TLS/SRTP and DNS, a verified tenant domain and supported network paths. The SBC terminates/normalizes the carrier or PBX side and connects it to Teams; it is a security and availability boundary, not merely a route entry.

The outbound decision chain is:

dialed string → tenant/user dial plan normalization → E.164 number → online voice routing policy → ordered PSTN usages → matching voice route → gateway/SBC → carrier.

A dial plan changes what number the user meant. A voice route chooses where a normalized call goes. A translation rule at a gateway can adapt the called/calling number for the SBC/carrier. Do not use all three layers to repair the same malformed number without documenting ownership.

Worked example 2 — Usage order can beat a more specific route. In a synthetic routing design, usage Local contains pattern ^\+1425555[0-9]{4}$ and usage Broad contains ^\+[1-9][0-9]{7,14}$. Both match +14255550123. Ordering Broad first sends this call through Broad; placing Local first selects Local. This is not automatic longest-prefix routing. Gateway order within one route is randomized, so a two-SBC list does not define deterministic primary/secondary preference. Use separate route priorities where needed, and test reachable-but-rejecting gateways as well as outages. Do not treat Calling Plan fallback as proof every Direct Routing failure is recoverable.

Plan redundant SBCs, routes and carriers with explicit priority/capacity and test failure—not only configuration presence. Local Media Optimization (LMO), Location-Based Routing (LBR) and media bypass solve different topology/regulatory/media-path problems. On-network conferencing can route applicable dial-out calls through Direct Routing. SBA guidance assumes the branch LAN and SBC/PSTN path still work, requires media bypass and a previously assigned survivability policy, and excludes VDI/web clients. Check the actual appliance-mode banner. Current supported functions include local-user VoIP behind the same SBA and selected forwarding/transfer/queue redirects; this does not make the cloud queue service available locally. Validate supported client/vendor versions, notification/recording limitations and failback after active outgoing calls finish.

Confirmed — VERIFY CURRENT (October 28, 2026). SBA considerations now allow a desktop client to reconnect after restart or reboot using authentication cached before the outage; expired tokens can be supported up to seven days after expiry. This is conditional continuity, not fresh offline authentication. The client still needs a prior connection to that SBA within 24 hours, and the CAE-token limitation remains roughly 30 minutes. Dynamic E911 location sharing and emergency-call-routing policies remain unsupported in SBA mode; emergency dial strings need matching regular voice routes. Validate the exact client, cache state, vendor build and approved emergency test plan.

Proposed practice extension — not executed: Extend Lab 6 tabletop with warm cache, absent cache, restart, expired token, no prior SBA connection and CAE cases. Predict which condition prevents service; run only vendor/carrier-approved nonproduction tests.

Answer checkpoint: Does reboot automatically invalidate SBA continuity? No; current docs support reconnect with pre-cached authentication, subject to the prior-SBA and supported-authentication-period limits.

Certified compliance recording and contact-center solutions integrate through supported Teams models. Assign policies and validate vendor responsibilities, recording notification/storage/access, agent routing and failure mode. The Queues app provides supported delegated call-queue/auto-attendant management and operational experiences under its licensing and policy model. Voice-application licensing requires Teams Premium for Queues app users; its delegated controls still need authorization. Calling Plan resource-account outbound transfers/callbacks/on-behalf-of calling need current PAYG/credits/overage funding rather than an ordinary minute bundle. Confirm Operator Connect carrier support separately; Direct Routing needs its own voice route.

Related item: A carrier says whether it accepted or delivered a PSTN call; Teams reports policy and cloud call legs; the SBC exposes SIP/TLS/media behavior. A complete Direct Routing incident often needs evidence from all three.

Choose a certified contact-center integration¶

Native auto attendants and queues can be enough for a reception desk. A customer-service operation needing business-system context, advanced routing and agent workflows should evaluate a certified contact-center integration. Compare the required workflow and agent endpoint before selecting the integration model: Connect joins the contact center through certified SBCs/Direct Routing; Extend integrates the Teams client and APIs; Unify uses Azure Communication Services to build the provider application on Teams calling infrastructure.

Original decision exercise — proposed, not executed: Compare a two-agent reception queue with a support desk needing customer records, skills-based distribution and contextual handoff. Write the requirement that native queues cannot satisfy, identify a currently certified provider/model, and rehearse customer-to-agent-to-Teams-expert transfer plus reporting and failure ownership. Certification does not prove every feature or license in your design; verify the provider's exact supported configuration. Escalate integration incidents to the contact-center partner first, which involves Microsoft when needed.

Teams Rooms and certified-device design¶

Start with room purpose, size, layout, acoustics, sightlines, lighting, accessibility, occupancy, join patterns, support model and budget. Then choose certified compute, camera, microphone, speaker, display, console/panel and cabling. A technically certified kit can still fail in a room with poor acoustics or obstructed viewing.

Teams Rooms Basic and Pro differ in entitlement and management features. Windows and Android rooms differ in hardware architecture, enrollment/management and feature timing. Rooms licensing limits Basic to 25 systems per organization, excludes panels, and requires Pro for multiple room systems using one account. Shared Space licensing is not a substitute for a Rooms license. For 28 single-system rooms with no advanced requirements, at least 3 need Pro because of the Basic cap; security, management or interop requirements can make more rooms need Pro. Plan a distinct room resource account, Exchange room mailbox/calendar processing, supported Rooms license, network access, Conditional Access posture, naming/location metadata and ownership.

Interoperability choices include:

Direct Guest Join for supported third-party meeting services;
Cloud Video Interop (CVI) for supported legacy/video conferencing endpoints joining Teams meetings;
SIP Guest Join extends a CVI provider’s supported legacy SIP-room scenario to external Teams meetings;
cross-platform SIP/H.323 dialing lets Windows and Android Teams Rooms with Pro, a supporting provider plan and the assigned video-teleconferencing policy call other SIP/H.323 endpoints;
BYOD rooms, where a user's laptop provides compute and the shared room peripherals/identity/location supply the space experience.

The third-party join comparison distinguishes DGJ’s browser path from SIP interop: examples include up-to-720p versus up-to-1080p video, one versus two displays, and HDMI sending support on the SIP path. These are conditional capabilities, not quality guarantees. Preserve external invitation bodies and verify mailbox processing and rewritten URLs when the join button is missing. Test the exact partner meeting type; a CVI agreement is not evidence every external feature works.

Personal/shared phones, displays, panels, Surface Hub, common-area phones, analog/SIP devices through SIP Gateway, bookable desks and full Teams Rooms are distinct endpoint types. Choose license, enrollment, configuration profile, calling/hotline behavior and management portal by device role.

Network and media design¶

Prepare the network for Teams before procurement or migration. Model personas, sites, concurrent modalities, direction and headroom. Use Network planner for assumptions and the current Network Assessment Tool for measured relay/connectivity diagnostics. The latter’s 1.9.0.0 listing is current as of this review; it is not a load/stress generator and was not executed here.

Favor local Internet breakout and a short supported media path. Define VPN split tunneling where appropriate. QoS requires aligned client port ranges, DSCP marking and network trust/queuing; it prioritizes during congestion but cannot add bandwidth. QoS guidance separates client source-port marking from service destination requirements; browser ports and Mac/mobile sharing markings differ from the default Windows table. Network roaming controls IP video and total average audio/video/video-sharing media bitrate at a site on supported Windows/macOS clients. Non-voice-enabled users need meeting-policy network lookup enabled. MBR is not a physical circuit reservation or a per-room capacity guarantee.

Upload tenant data and reporting labels to CQD so subnet/building/site patterns become meaningful. Network roaming policies can apply media settings based on network location. Maintain topology data as subnets change. CQD building labels, network policy sites and LIS emergency locations are separate datasets; updating one does not populate the others. eCDN reduces duplicate video delivery across constrained corporate networks for applicable town hall/streaming scenarios; validate provider, licensing, topology, capacity and test-event telemetry.

Worked example 3 — Measure eCDN offload. Assume 240 viewers each demand 3 Mbps: unoptimized site delivery is 720 Mbps. A rehearsal measures 540 Mbps served through peers and 180 Mbps from the WAN: offload is 75%. With 120 Mbps of other WAN traffic, total is 300 Mbps. A 350 Mbps link with a chosen 20% reserve permits 280 Mbps, so this sample still misses the target by 20 Mbps. Microsoft eCDN reduces duplicate delivery; it does not remove presenter contribution, Wi-Fi or peer-network constraints. Measure again with the real audience/network layout.

Metric	Interpretation in context
Packet loss	Missing audio/video data; inspect direction and burst pattern
Jitter	Variable packet arrival; can exhaust playout buffers
Round-trip time	Conversational delay and talk-over risk
Bitrate/resolution/frame rate	Adaptation to bandwidth, policy, CPU or endpoint constraint
Concealment/poor-stream percentage	User-impact signal that needs endpoint/path correlation

2. Configure and manage Teams meetings, webinars, and town halls¶

Meeting policy stack and advanced experiences¶

Configure meeting settings first for tenant-wide behavior, then policies for cohorts, templates/labels for repeatable protected scenarios and meeting options for a specific event. For ordinary supported Teams policy types, direct assignment beats ranked group assignment, then global; lower rank numbers win and nested membership does not inherit. Emergency site policy has a different precedence rule described below. Inspect the effective result for each policy type. Allow propagation time, but do not use propagation as an explanation without evidence.

Build a test matrix for organizer, co-organizer, presenter, internal attendee, guest/external attendee and anonymous participant. Validate scheduling, lobby, who can present, content sharing, chat, reactions, apps, recording/transcription, attendance reports, breakout rooms, Q&A and captions. Protected meetings may add watermarking, encryption, recording restrictions or sensitivity-label enforcement under applicable licensing.

Controlled-content scenarios require more than disabling screen sharing: decide who can present, whether attendees can request control, which apps/polls/forms are permitted, how files are shared and whether external participants can access downstream artifacts. Recording and transcript retention/access are governed by storage location, meeting roles, sharing and Purview policies as well as the Teams meeting setting.

Virtual appointments add external scheduling/join and operational requirements. Determine whether basic scheduling is enough or advanced appointment policies, queue views, notifications, analytics and branding are required. Test a browser/mobile external attendee rather than only a signed-in employee.

Create and assign policy packages¶

A Teams policy package bundles policies for a role. In Policy packages > Add, name a custom package, describe its purpose, select policy types and named policies, and save. Each user targeted by a custom package requires Teams Premium; check entitlement before assignment. Packages can target individuals, groups or batches.

Original exercise — proposed, not executed: Create a pilot Reception package containing chosen meeting and calling policies. Record each user's prior assignments, assign the package to a test cohort, and inspect the effective policy of each type after propagation. A successful package assignment alone does not establish the resulting meeting/calling behavior: check direct assignments and ranked groups using the policy precedence above, then test a meeting and call. For rollback, restore the recorded assignments and verify their effective results. Editing a policy in the package affects its assigned users, so pilot a change before wider rollout.

Audio Conferencing¶

Assign the entitlement to organizers who need to create dial-in meetings. Configure the default conference bridge language and numbers, acquire toll/toll-free inventory, determine which numbers appear in invitations and configure user-specific default number/toll-free behavior where supported. Validate dial-in, meeting ID/passcode behavior, dial-out/call-me, anonymous restrictions and cost controls in every required country.

Changing the default bridge does not necessarily rewrite existing meeting invitations. Define when organizers must resend or recreate meetings and test a previously scheduled meeting during migration.

Webinars and town halls¶

The unified Events experience preserves webinar/town-hall policy controls. Optimize for large audience chooses town-hall behavior and is mandatory above 1,000 attendees; disabling it uses webinar controls. Retain the exam’s format distinctions while checking actual interaction and license requirements. Legacy live events retired June 30, 2026, with a grace period through February 28, 2027 for previously scheduled events.

For webinars, configure who may schedule, organization-only versus public registration, presenter/co-organizer roles, registration capacity/questions, communications, interaction, recording/publishing and reporting. Teams Premium may add advanced webinar functions; verify current license and organizer requirements.

For town halls, configure scheduling policy, organizers/co-organizers, presenters, attendee access, Q&A/chat/reactions as supported, recording and reporting. Plan rehearsals, producer/presenter device and network readiness, support bridge and contingency content. For corporate networks, evaluate and configure eCDN and validate with a representative audience; do not first test distribution at the production event.

Related item: Webinar registration controls who receives an event join experience; lobby policy controls admission to a meeting. Similar-looking attendee problems can belong to different control planes.

Town halls across Teams, Viva Engage and recorded video¶

The objective's Microsoft Stream wording needs a current workflow check. Viva Engage town-hall instructions describe a community administrator creating an event from the community's Events or Conversations tab, selecting the Teams town-hall broadcast option, entering schedule/details, checking public/private audience, assigning presenters/co-organizers and choosing Teams production or RTMP-in before publishing. Use a test event for rehearsal; its setting cannot be changed after creation. Teams supplies production and recording, while Engage supplies community discovery and discussion. Legacy live-event instructions must not be assumed to describe the replacement town-hall workflow.

For the video side, the current town-hall recording instructions use Calendar > past event > Manage event > Recordings > Publish, with Publish from OneDrive when applicable. Current playback/editing UI can say Clipchamp where older material says Stream. Original check — proposed, not executed: rehearse the Engage event with producer and attendee accounts, publish an approved recording, then verify the attendee's actual access and expiry. Separately check the recording file permissions; a shared event or video URL alone does not grant access.

Configure avatars with an expiry boundary¶

VERIFY CURRENT: Avatars administration announces retirement on April 2, 2027, with support until then. Treat this as bounded exam practice with a camera/audio fallback.

In Teams apps > Manage apps, inspect Avatars, allow it for a small test audience using the tenant’s current app-management model, and verify the relevant service plan, supported client/hardware and network endpoints. Have the pilot user install/configure an avatar and join a synthetic meeting; compare an excluded persona. Pinning alone does not grant availability. The article mixes manual-install and setup-policy guidance, so record the actual installation path instead of assuming an app policy installed it automatically.

Original exercise — proposed, not executed: Record audience, entitlement, installation, client, meeting outcome and denied case separately; restore the prior app configuration afterward. Answer checkpoint: A visible app icon does not establish access or successful avatar use.

Evaluate Teams Premium in the admin center¶

In Teams admin center > Teams Premium, use Overview, Licenses, Usage reports and Feature advisor to compare a requirement with entitlement, configuration and measured use. The dashboard guidance notes that some former Premium features moved to Teams Enterprise on April 1, 2026; verify the exact capability before choosing an add-on.

VERIFY CURRENT: an eligible discovery trial is started through Try for free and currently offers 25 licenses for 30 days. It is unavailable after a previous trial or with a paid Premium subscription. The dashboard’s usage view covers the preceding 30 days; it is not an immediate acceptance test.

Original exercise — proposed, not executed: Tabletop a five-user evaluation: name one required feature, assign an owner, record license/configuration prerequisites, test allowed/denied behavior and define success and expiry cleanup. Inspect feature-advisor suggestions without automatically enabling them. Answer checkpoint: Licenses assigned or a feature marked configured do not prove the pilot’s intended experience.

3. Implement and configure Teams Phone¶

User calling configuration and policies¶

For a voice user, verify identity, usage location, Teams Phone entitlement, PSTN-connectivity entitlement/model, TeamsOnly requirement where applicable, number assignment, emergency location/policy, dial plan, voice routing policy and calling/caller-ID/voicemail policies. Test internal Teams call, inbound PSTN, outbound local/national/international, transfer, forwarding, voicemail and emergency validation using approved procedures.

Policies have separate jobs:

Object	Controls
Calling policy	Call features such as forwarding, delegation, simultaneous ring, call park, private lines and inbound handling where applicable
Caller ID policy	Presented caller identity and replacement/service-number behavior
Tenant dial plan	Number normalization rules
Online voice routing policy	PSTN usage/routes for Direct Routing
Emergency calling policy	Emergency notifications and related user experience
Emergency call routing policy	Emergency numbers, routes and dynamic behavior
IP phone/configuration profile	Supported phone/device settings
Voice applications policy	Delegated auto-attendant/call-queue/Queues app capabilities

Call park retains a call against a retrieval code; group call pickup, simultaneous ring, forwarding and delegation distribute or hand off a user's calls. Cloud Voicemail handles unanswered calls with policies/settings for prompts, transcription and routing. Outbound restrictions and PSTN-usage design limit cost/risk. Unassigned-number routing sends calls to numbers in specified ranges to an announcement, user or voice application instead of failing unpredictably.

Worked phone-policy matrix¶

Original example — proposed; service/tenant execution not performed.

For every pilot, create or select the specific policy, assign it to the test account where supported, inspect effective assignment and then test the actual feature. VERIFY CURRENT: interface paths, number models, licensing and propagation.

Requirement	Configuration and decisive test
Park and retrieve a call	Enable call park in its policy, choose an unused pickup-code range and timeout; park, retrieve by code, then verify unanswered timeout returns the call
Shared-phone interface	Use an IP Phone policy sign-in mode matching the entitled device purpose; sign out/in after propagation and compare the actual common-area/meeting/user UI
Mobile incoming-call preference	Set mobility policy to Teams, native dialer or user-controlled; verify the SIM-enabled device's actual first-ringing experience
Custom hold audio	Configure the call-hold policy and enable Music on hold in calling policy; a file upload alone is insufficient
Inbound nuisance numbers	Use tenant-global PSTN blocking patterns and exemptions with Teams PowerShell, then test a matching and exempt number; this is not per-user caller-ID hiding and currently lacks TAC management
Required recording	Provision the certified partner recording application and policy, assign the intended scope and verify notice, capture, failure and retention with compliance owners; convenience recording is a different control

A dial-plan change cannot repair a blocked feature or missing recording integration. Use synthetic calls; emergency calling and SBA paths have separate recording limitations and must not be assumed covered.

Question: A hold-audio file is configured but callers hear none. Which dependent control matters?

Answer: Music on hold must also be enabled in the calling policy; verify effective assignment and an actual test call.

Teams settings and policies reference · Set Microsoft Teams Phone device user interface · Block inbound calls · Microsoft Teams compliance recording (third-party)

Worked calling features — recipient and timing¶

Original example — proposed; service/tenant execution not performed.

Distinguish the policy that permits a feature from the user's call-handling configuration. In a disposable voice-enabled pilot, first verify the relevant calling-policy controls, then configure the target user:

Need	Setting and expected test
Another colleague answers after no answer	Configure a call group and unanswered handling; verify membership, ringing order and timeout
Immediately send calls elsewhere	Configure forwarding destination; a primary-line test reaches that destination instead of ringing the original user
Ring both user and colleague	Configure simultaneous ringing; either endpoint can answer the same incoming call
Assistant acts for manager	Add an explicit delegate and permitted make/receive/manage rights; compare inbound handling and outbound on-behalf-of identity
Second confidential inbound number	Assign one private line matching the primary number type; verify special indication and private-line handling rather than assuming forwarding or delegation applies

Private lines are inbound-only and ignore ordinary forwarding, simultaneous-ring and call-group handling; Do not disturb and voicemail still matter. A private line is not a second outbound identity. Turning off delegation policy also requires cleaning up existing delegation relationships. VERIFY CURRENT: client support, licensing and number assignment. Record before/after Get-CsUserCallingSettings output where authorized and restore the pilot settings.

Question: Will a manager’s ordinary forwarding rule necessarily forward a private-line call to the assistant?

Answer: No. Private-line calls bypass those ordinary inbound-handling rules; test private-line voicemail and DND separately.

Calling policies in Teams · Configure call forwarding and delegation settings · Configure private lines in Microsoft Teams

Assign phone numbers in a controlled PowerShell batch¶

Original exercise — proposed, not executed. VERIFY CURRENT: Prepare a CSV with Identity, TelephoneNumber, NumberType and LocationId for two authorized lab users. Validate unique users/numbers, licenses, inventory, location and number model; save existing assignments before mutation. Render the planned rows for review. For each approved row, the documented operation is:

Set-CsPhoneNumberAssignment -Identity $row.Identity -TelephoneNumber $row.TelephoneNumber -NumberType $row.NumberType -LocationId $row.LocationId -ErrorAction Stop

Wrap the per-row operation in try/catch and retain success/error separately. Query each number with Get-CsPhoneNumberAssignment, compare AssignedPstnTargetId and location with the intended account, then test ordinary inbound/outbound calling. Plan restoration from saved state; do not blindly retry a partly successful batch. An on-premises synchronized number can prevent online assignment. Inventory queries default to 500 rows; use documented pagination when checking larger sets.

Set assignment contract · Read assignment and pagination

Answer checkpoint: Two successful requests are not proof that both users can call.

Auto attendants and call queues¶

An auto attendant applies greetings, menus, schedules and routing. A call queue distributes callers to agents. Design them on paper before configuration:

create the application and its resource account where needed (directly answering applications require one; nested applications can use the parent’s account);
assign the Microsoft Teams Phone Resource Account license;
assign a service number if PSTN callers must reach it;
configure language, greetings/music, business hours and holidays;
configure destinations, agents, routing method, presence/opt-in, callback, overflow and timeout;
assign authorized users/voice applications policy for delegated management or Queues app;
test every branch, including closed, holiday, no agent, overflow, timeout and transfer failure.

Resource-account guidance distinguishes the identity/number anchor from the application. Do not enable voice-application resource accounts for interactive sign-in. Resource accounts for Rooms are a different device identity type.

Routing methods and call priorities should match work rather than preference: serial, round robin, longest idle or attendant-style behavior can create different fairness and pickup outcomes. Presence-based routing, agent opt-out and callback change capacity assumptions. Queue configuration makes callback eligibility, completion of music, caller response and agent answer share the timeout budget. A callback request can itself time out. Validate E.164 caller eligibility and failure notifications. Conference mode is recommended but incompatible with calls arriving from an LBR-enabled Direct Routing gateway; transfer mode is legacy, so retain a support-plan follow-up. Use custom music/prompts only with appropriate rights and a fallback.

Configure queue outbound identity¶

Distinguish a user's caller-ID substitution policy from an agent deliberately calling on behalf of a queue. In Voice > Call queues, edit the queue and use Assign calling ID > Add to select the intended resource account, then save. A pilot agent chooses that queue identity in Calls. Separately, Voice > Caller ID policies controls ordinary user-number replacement and user override. See queue setup and caller-ID policies.

VERIFY CURRENT (October 28, 2026): The setup page links the November 1, 2025 licensing change. Its broad Calling Plan category must be read with the dedicated licensing reference: funded PAYG/credits or applicable overage is required for Microsoft-number on-behalf-of calls. A minute bundle alone does not establish that funding. Verify the actual carrier path; Direct Routing requires the resource account's routing policy, and Operator Connect has its own number requirements.

Original check — proposed, not executed: Make approved synthetic calls as self and as queue, record the selected identity, number received and call success, then restore the pilot assignment. Correct display alone does not prove outbound entitlement, routing or return-call behavior; carrier name display can differ.

Dynamic emergency calling¶

Treat emergency calling as a life-safety design. Define civic addresses/emergency locations, network sites/subnets/switches/ports/wireless access points, Location Information Service (LIS), trusted IPs, emergency numbers, routing policies, notification/security-desk workflow and carrier/PSTN responsibility. A verified address is a prerequisite for applicable number assignment; it does not prove that a roaming client receives the right dynamic location.

Worked example 4 — Location matching must use the client’s prefix. A synthetic client reports 10.20.30.170/25; its network ID is 10.20.30.128/25. An LIS entry only for 10.20.30.0/24 does not satisfy that documented exact network-ID calculation. If a mapped wireless access point identifies Floor 3 while the subnet says Floor 1, the WAP wins. Dynamic emergency guidance orders WAP, switch/port, switch, then subnet. Confirm the client’s actual reported location after propagation; successful number assignment does not validate LIS.

Network-site emergency policy overrides the user’s policy at that site. Keep ordinary Direct Routing emergency-route requirements separate from the special Shared Calling procedure: Shared Calling requires a routing policy for every number model, with PSTN usages only for its Direct Routing path. Its emergency callback numbers must be routable, of the right number type/country and not reused across policies. A toll-free shared number cannot supply the empty-list callback fallback.

Test only through carrier-approved validation numbers/processes. Validate fixed office, remote user, VPN, branch/SBA, room/common phone and mobile scenarios. Record dispatchable-location behavior, callback, notification, route and failure escalation. Revalidate after network, carrier, SBC, office or number changes. Dynamic emergency calling/security notifications are not supported by the Teams web client. Direct Routing requires appropriate PIDF/LO support plus the emergency routing provider/ELIN design. Treat emergency-route tests and ordinary PSTN tests as separate acceptance results.

Related item: A dial plan normalizes an ordinary number. Emergency call routing recognizes emergency dial strings and selects emergency behavior/routes. Never hide an emergency-design defect inside a broad normalization rule.

Direct Routing implementation and troubleshooting¶

Implement in dependency order: verified domain/DNS and public certificate; supported SBC software/configuration; network/firewall; SBC pairing and health; PSTN usages/routes/gateways; voice routing/dial plans/translations; user assignment; emergency/LBR/LMO/media-bypass/SBA features; validation and monitoring.

When a call fails, collect UTC time, calling/called number, user, policy, correlation/call ID, location/client, SBC SIP ladder and carrier reference. Trace:

did Teams normalize the number as intended?
did effective voice routing policy expose a PSTN usage?
did an enabled route match the number and select a healthy gateway?
did TLS/SIP reach the SBC with expected identities/numbers?
did SBC translation/routing send it to the carrier/PBX?
what final SIP response and reason came from each boundary?
if signaling succeeded, did media negotiate and flow both ways?

Current media-protocol guidance still limits the SBC ICE-Lite candidate to IPv4; signaling IPv6 support does not establish IPv6 media-bypass support. Correlate the selected endpoint after forked early responses, final answer and ICE nomination. One-way audio usually points to media path/NAT/firewall/SBC negotiation, not dial-plan matching. A 4xx/5xx/6xx SIP response must be interpreted at the hop that generated it. Use synthetic tests and Call Analytics/CQD trends, then prove failover by disabling a safe test route/gateway rather than waiting for an outage.

Worked example 5 — Transfer acceptance is not destination answer. In a synthetic Direct Routing trace, a blind transfer receives 202 Accepted at 12:00:02 and the original leg ends immediately; the target later returns busy at 12:00:05. June 2026 transfer behavior permits the original-leg termination after SBC acceptance. Diagnose the subsequent target leg instead of treating that BYE alone as premature. Consultative transfers still require their own progress/completion handling. Preserve the Refer-To/Referred-By strings and correlate SIP legs; do not rewrite them as ordinary dial strings.

Configure on-network Audio Conferencing¶

On-network Conferencing uses Direct Routing for supported meeting calls between the organization's telephony network and Microsoft Audio Conferencing. It is not a supported way to add non-Microsoft dial-in numbers or send bridge calls onward to the PSTN. It must not be deployed with telephony equipment in India; pay-per-minute Audio Conferencing is excluded.

After Audio Conferencing licensing/service and SBC readiness, route on-premises calls to your Microsoft bridge numbers through the SBC/PBX. For meeting dial-out to on-network destinations, create PSTN usages and matching voice routes, create an OnlineAudioConferencingRoutingPolicy using those usages, and grant it to the meeting organizer. The organizer's policy controls routing even when another attendee starts dial-out; a user's ordinary voice-routing policy is a separate object. Configure the PBX/SBC destination route and optional extension normalization.

Original tabletop — not executed: Organizer A has the pilot policy; attendee B initiates a call to an approved internal extension. Trace A's usage, matching route and SBC destination, then compare a nonmatching destination. Record expected fallback and prevent unintended PSTN routing. Validate only with authorized lab numbers and retain the SBC and meeting evidence.

4. Configure and manage Teams Rooms and devices¶

Resource accounts, enrollment and secure access¶

Each Teams Room needs its own room resource account/mailbox and supported license. Configure calendar processing, booking behavior, display name/location/capacity and Exchange room-list/Places metadata. Use a naming standard that identifies site/building/floor/room without exposing sensitive details.

Microsoft's room-account guidance and Conditional Access guidance require a device-compatible identity design. Do not apply ordinary interactive MFA, smart-card or certificate prompts to a headless shared-room sign-in. Use supported modern authentication, compliant device, known network/location and scoped Conditional Access. Pilot on test rooms and keep an exclusion/recovery path while proving policy behavior.

Android Teams devices use current AOSP/Microsoft Device Ecosystem Platform (MDEP) enrollment paths and supported Intune policies. Windows Rooms use Windows/Intune enrollment and device-specific configuration. Enrollment is not merely inventory: it supplies device compliance and access context. AOSP enrollment uses one corporate-owned, user-associated enrollment profile with For Microsoft Teams devices enabled. Teams devices use account credentials rather than its QR code. An expired token blocks enrollment/sign-in. Intune entitlement and assigned AOSP compliance policy must be ready before compliance-based CA; Basic alone does not include Intune. MDEP/OEM firmware readiness and Intune enrollment are different checks. Scope Teams compliance by enrollment profile when other AOSP devices share the tenant.

The device-code policy guidance uses persistent, monitored exceptions for real device accounts and excludes Device Registration Service from the blocking policy. Exceptions are account-wide, not limited to a single room action. Include reprovisioning/recovery even for passwordless devices; confirm the installed authentication design rather than copying a legacy ROPC procedure.

Operations and lifecycle¶

The Android management transition makes Rooms Pro Management the primary portal, with overlapping TAC workflows retiring during September 2026. Devices appear automatically after prerequisites; configuration profiles must be imported as settings templates. Portal entitlement does not mean every device needs a new Pro license. Check access/RBAC, required endpoints, minimum Admin Agent and actual device actions; an Entra AU-scoped administrator needs appropriate PMP role delegation. Existing pause settings generally carry over, but mandatory migration Agent updates are non-pausable. Follow the dated transition guidance and cloud-specific release status before relying on a rollout date. Device management guidance also distinguishes partial settings application from overwriting a full profile. Compare effective device settings after template import. The XML reference remains relevant to the exam, but Microsoft now recommends PMP settings and no longer publishes new XML settings. A correctly processed SkypeSettings.xml is consumed and deleted at startup; malformed XML can apply earlier settings before failure, omitted values stay unchanged, and unknown elements are ignored. File disappearance alone does not prove the intended outcome. Validate supported elements/case, keep a known-good configuration, and inspect effective settings after restart.

Create update rings: pilot representative hardware/peripherals first, observe meetings/sign-in/HDMI/content/camera/audio, then stage production. Current Rooms release notes separate installed app updates from web-delivered features. Windows 5.6.210.0 (August 3) enables the Windows 11 25H2 update; an app update can therefore precede an OS change. September Android notes include SIP/H.323 dialing, subject to supported partner/configuration. Admin Agent 865 lists provisioning/health fixes, with government-cloud rollout separate from public cloud. Match OEM, app, OS, Agent, cloud and management surface in acceptance evidence. A device shown as online is not proof that camera framing, microphone pickup, speaker output, dual display, HDMI ingest or guest join works.

September Android build 1449/1.0.96.2026249711 also fixes continued EWS calendar use on certain devices. The release excludes Cisco pending a later build. Verify the model-specific fix before disabling Exchange Web Services; a release-note date does not prove the device has migrated.

For Windows Rooms, exclude unsupported domain Group Policy/user security settings that interfere with the appliance account while retaining required device security. Validate custom display layouts and XML syntax/version support. For Android, use configuration profiles for supported settings, IP phone policies, hotline and remote deployment, and verify that the correct account/device type receives them.

SIP Gateway brings supported SIP devices into Teams. It requires a PSTN-enabled phone number, does not publish device presence for presence-based queue routing, and can offboard disconnected signed-in devices after 30 days (signed-out devices after 14). A desk phone in storage may need reprovisioning; basic SIP connectivity does not prove native Teams feature parity. Verify device model/firmware, network provisioning, authentication, policy, calling/emergency behavior and feature limitations. Common-area and conference phones, panels and displays need the correct shared-device identity/license and operational profile, not a reused personal account.

Configure Windows room display branding¶

Windows Rooms custom backgrounds require Teams Rooms Pro for the enhanced method; Basic does not support custom backgrounds. Prepare distinct, legible images for the main display and, when enabled, the extended display. Copy the files to the documented Rooms LocalState folder, then deploy a supported SkypeSettings.xml fragment:

<SkypeSettings>
  <Theming>
    <ThemeName>Custom</ThemeName>
    <CustomBackgroundMainFoRDisplay>Pilot-main.png</CustomBackgroundMainFoRDisplay>
    <CustomBackgroundExtendedFoRDisplay>Pilot-second.png</CustomBackgroundExtendedFoRDisplay>
  </Theming>
</SkypeSettings>

Original check — proposed, not executed: With dual-display mode already configured, confirm both files, resolution/aspect-ratio requirements and entitlement; apply during a maintenance window and inspect each actual display, calendar readability and restart recovery. Missing required images or undersized files can produce No Theme on all displays. Keep the prior XML/images for rollback. The console image is optional. Use the current enhanced template rather than assuming that a legacy panoramic image or an Android console image has the correct dimensions. File processing alone does not prove the visual result.

Configure Android and panel display branding¶

Separate the Android room's main/extended/console images from a panel's single home-screen image. VERIFY CURRENT (October 28, 2026): The Android background article requires a supported app and Rooms Pro, and now directs administrators to Pro Management > Rooms > room > Settings > Theme > custom wallpaper. The no-custom-background statement in the remote-settings reference belongs to its Windows tab; it does not contradict this Android-specific workflow. Inspect the actual Android tenant/device controls and current app prerequisites before rollout, and preserve the prior profile/background for recovery.

For a supported Teams panel, use Teams devices > Panels > Configuration profiles > Add/Edit > Device settings > Background > Use a custom background, upload one image, save and assign the profile to a licensed pilot panel. Check the current management transition before using TAC. A 200 KB JPG at least 1280×720 meets both the panel article's stated range and the generic profile page's stricter under-500-KB note.

Original check — proposed, not executed: Mark each image with its intended screen; inspect cropping, contrast, room/calendar visibility and post-restart state. Change only the pilot, record the applied profile/settings and prove rollback to a built-in image. An uploaded asset is not an applied display configuration.

Keep workstation GPOs off room appliances¶

Treat Windows Rooms as appliances rather than ordinary employee PCs. The Rooms security guidance says the local Skype kiosk account must not be edited or given a password; a generic local-account password policy can break automatic sign-in. That local account is different from the cloud room resource account and the administrative recovery account.

The Windows Update device guidance also says Rooms need no additional update policies and warns that offering policies can conflict with its validated-update management. Review the room computer's GPO/MDM scope: exclude workstation password-hardening and update-offering settings that override these appliance requirements, while retaining supported security controls.

Original inspection exercise — proposed, not executed: Compare a working room's effective policy with a pilot after a workstation GPO is linked. Identify the setting attempting to change the Skype account or update channel, narrow the policy scope, then verify automatic room sign-in, approved update status and recovery access after restart. Do not disable domain security globally or interpret an exclusion as permission to stop patching. Record the actual applied policy and supported exception.

Provision a supported SIP handset¶

SIP Gateway configuration requires a compatible model/firmware, approved network access and an effective calling policy with SIP devices can be used for calls enabled. Reset a disposable pilot device using its OEM procedure. Supply the documented regional provisioning URL through the vendor's supported DHCP option or device web utility; Americas uses http://noam.ipp.sdg.teams.microsoft.com. The bootstrap URL uses HTTP, not an invented HTTPS replacement. DHCP options vary by vendor.

Original provisioning check — proposed, not executed: After onboarding shows Teams sign-in, open Teams devices > SIP devices > Actions > Provision devices, add the pilot's MAC/location and generate its verification code. Dial *55* followed by that code on the correct physical handset, then complete the documented user-pairing/sign-in step. Onboarding, tenant enrollment and user sign-in are separate states. Verify the intended identity plus approved inbound/outbound and location tests before releasing the device; never use an emergency number as an ad hoc test. Record firmware, effective policy and network evidence. A missing logo suggests bootstrap/network/firmware trouble; a logo without user service needs sign-in/policy investigation.

Create and manage Teams device tags¶

Device tags organize device inventory. With a Teams Device Administrator or Teams Service Administrator role, the documented TAC route is Teams Devices > device type > Actions > All device tags. Create Pilot-East, then select test devices, use Manage tags, and apply it. VERIFY CURRENT: management is moving to Pro Management; confirm the available portal workflow before following older TAC steps.

Original exercise — proposed, not executed: Filter one device type by Pilot-East, compare match all with match any using a second tag, and inspect the selection before a bulk action. Tags do not automatically assign configurations or enforce policy. They follow the signed-in resource account, so moving that account can move the apparent device grouping. Rename the tag, verify the filtered inventory, then remove its assignments and delete it. Keep room/location metadata and team-member @mention tags separate from this device-management grouping.

Enable and verify hotline on a shared phone¶

A hotline/PLAR phone automatically calls a configured directory contact or PSTN destination when its handset is lifted. The common-area phone guide separates this from ordinary advanced-calling features.

VERIFY CURRENT: use a supported certified phone, current app, common-area sign-in policy and appropriate shared-device entitlement. For the documented TAC route, select Teams devices > Phones > Configuration profiles, open the profile assigned to the pilot phone, and enable hotline under Call settings. Configure the intended display name and destination through the supported device/profile workflow. Recheck Pro Management transition and installed firmware rather than applying a legacy path blindly.

Original exercise — proposed, not executed: Use an authorized non-emergency lab destination. Lift the handset and prove the correct endpoint rings; compare a wrong/unreachable destination and document recovery access. Disable the test setting afterward and confirm normal behavior returns. Answer checkpoint: An enabled toggle without a correct destination and successful call does not establish a working hotline.

Room features and flexible workspaces¶

Optional room capabilities include HDMI ingest, content camera, casting, proximity join, room remote, speaker recognition/intelligent audio/video, Direct Guest Join and hot desking. For each, record hardware/room/license/client prerequisites, privacy/accessibility effect and support fallback.

For BYOD rooms and bookable desks, define the physical-space inventory, peripheral association/discovery, building/floor/room metadata, booking/resource behavior and user workflow. Import or discover devices and assign them to the correct space; monitor adoption and bad mappings. A BYOD room may rely on a user's device for compute, but shared peripherals and space identity still require governance.

Worked example 6 — Desk discovery needs distinct signals. In bookable desks, assume an auto-association threshold of three unique users. Alice docks four times, Bob once and Carol once: six events supply three distinct users, so the threshold is reached only with Carol. Alice’s repetitions alone do not qualify. Auto-association is for individual desks; pools follow their own mapping/capacity path. Missing, duplicate or invalid peripheral serial numbers are unsupported. Verify the actual desk and peripheral association, allow directory/association propagation, and distinguish a single reserved seat from a pool’s remaining capacity. Work-location updates also depend on policy and user opt-in.

Worked optional-room acceptance matrix¶

Original example — proposed; service/tenant execution not performed.

Treat each feature as a separate acceptance result; settings and platform/OEM support vary. VERIFY CURRENT in the device's current Pro Management settings and supported feature matrix.

Feature	Configure or inspect	Evidence and failure clue
HDMI ingest	Separate local display, Automatic screen sharing and HDMI audio-sharing settings	A local image is not proof remote participants received video/audio
Content camera	Select the actual content camera and enhancement/rotation settings; confirm room license and OEM support	Share a marked whiteboard with a remote participant; test framing, legibility and camera movement
Casting	Enable supported Bluetooth beaconing and select the intended room from the casting experience	Verify intended room and shared image; do not equate detection with content delivery
Proximity join	Configure Bluetooth or supported ultrasound beaconing and its automatic-accept control; select the nearby room during join	Validate correct-room discovery and echo-free meeting entry; isolate Bluetooth/client/network causes
Room remote	Enable Remote control from personal devices and Bluetooth beaconing; ultrasound alone does not enable room remote	Verify permitted meeting controls from the phone/client and loss-of-range/recovery behavior
Hot desking	Match shared-device license/account and IP Phone policy to actual hardware; AllowHotDesking enables it and HotDeskingIdleTimeoutInMinutes sets automatic logout	Sign in, end session and prove personal information is removed before the next user
AI voice/video	For supported framing, enable Cloud IntelliFrame; for biometric enrollment, scope CsTeamsAIPolicy EnrollVoice/EnrollFace/PassiveVoiceEnrollment and obtain user opt-in, then configure the separate room-recognition permission	Verify attribution/framing in a synthetic meeting, with privacy and an ordinary-audio/video fallback

The historical Teams displays article describes hot desking but also states that its app support ended September 3, 2025. Its two-year compatibility effort is not an extension of support. Do not procure or present that legacy display workflow as an unqualified current deployment recommendation; validate the supported shared-device category and installed app instead. No room feature was executed here.

Question: Does content visible on the front-of-room display prove it was shared remotely?

Answer: No. Check remote video and audio separately; HDMI automatic-sharing and audio settings can create different outcomes.

Settings available for remote management in Teams Rooms Pro management portal · Content cameras · Microsoft Teams displays · New-Cs Teams IPPhone Policy · Overview and configuration of voice and face enrollment · Teams Rooms (Windows and Android) home screen and admin controls

Create and publish room types and standards¶

Use the Standards and Room planner to turn a room-type design into a published planning standard. With the documented Teams Rooms Pro or Teams Shared Device entitlement, create a Room Standard, supply a name/description, choose the meeting-space type and recommended certified hardware from inventory, review, then publish to the appropriate region, site, building or room. Published standards support planning; they do not prove that hardware was installed or device policies applied.

Original exercise — proposed, not executed: define Small-BYOD-East with an agreed display, camera and audio design. Reconcile discovered/imported hardware, map devices to the intended space/type, publish the standard to one pilot building and compare the planned room with actual inventory and a test call. An unassigned peripheral or wrong location is an inventory/mapping failure even when the standard is published. Record the changes needed before wider publication.

Device troubleshooting¶

Start with resource-account sign-in test, license, password/credential lifecycle, Conditional Access sign-in logs, device compliance/enrollment, Exchange calendar processing and service health. Then inspect Teams admin center/Pro portal health, app/OS/firmware version, network/DNS/proxy/time, peripherals and logs.

For remote provisioning failure, validate supported device/firmware, management URL/DHCP/DNS, provisioning code validity and physical possession. For proximity/casting failure, isolate Bluetooth/network/client/room policy. For calendar failure, compare Exchange room mailbox behavior with Teams sign-in rather than resetting the device immediately. Preserve logs before factory reset.

Integrated scenarios¶

Scenario 1: Multinational Teams Phone migration¶

Segment countries and personas. Choose Calling Plan, Operator Connect, Teams Phone Mobile or Direct Routing per regulatory/carrier/feature need; inventory licenses and numbers; design dial plans, routing, emergency calling and support boundary. Pilot porting and test inbound/outbound/emergency/voicemail/failover. Use CQD and carrier/SBC evidence during phased migration rather than changing all users at once.

Identify room type, resource accounts, license, enrollment/compliance and exact Conditional Access result. Compare a working pilot room and inspect supported authentication requirements. Narrowly revise the device-account policy, validate sign-in plus blocked misuse, then expand through rings. Do not exempt all shared accounts or disable tenant MFA globally.

Scenario 3: Executive town hall with remote sites¶

Select town hall rather than a meeting from scale/interaction; assign organizers/presenters; configure policy, eCDN and recording/Q&A; validate producer devices and contribution networks. Run a representative rehearsal, monitor real-time/presenter evidence and provide alternate content/support. Afterward, correlate attendee analytics, site delivery and support tickets.

Hands-on labs¶

Execution boundary: This review performed offline calculations/decision checks and repository validation only. No tenant, room, SIP, emergency, PSTN, media, SMS, portal migration or network diagnostic was executed. Use a test tenant, lab numbers and synthetic data. Do not place test emergency calls except through carrier-approved procedures.

Lab 1 — Architecture and licensing matrix¶

Design two sites and four personas. Compare all PSTN models, user/resource/room licenses, number types, room/device options and support owners. Evidence: decision matrix, assumptions and rejected alternatives.

Lab 2 — Network, CQD and eCDN plan¶

Model concurrency and bandwidth, document ports/QoS/split tunnel/MBR, upload sample tenant topology/reporting labels and plan eCDN validation. Evidence: capacity worksheet and before/after test plan.

Lab 3 — Meetings, webinar and town hall controls¶

Create a meeting policy/template and event policies. Test organizer, presenter, internal/external/anonymous attendee, recording/transcription and protected content. Evidence: effective-policy matrix and experience results.

Lab 4 — Audio Conferencing and voice-user build¶

Map bridge/toll/toll-free and invitation behavior. Build a lab voice user with number, dial plan, calling/caller-ID/voicemail/emergency policies. Evidence: entitlement/object map and call test sheet.

Lab 5 — Auto attendant and queue¶

Create resource accounts, schedules/holidays, menu, queue, agents, callback/overflow/timeout and authorized-user policy. Test every route. Evidence: call-flow diagram and positive/failure recordings or logs.

Lab 6 — Direct Routing tabletop or lab¶

Build a sanitized topology with SBC, certificate/DNS, routes/usages/policies, translations, emergency and redundancy. If available, trace a synthetic call and failover. Evidence: SIP/routing trace and boundary-specific runbook.

Lab 7 — Teams Room deployment¶

Create a room account and booking policy, model supported Conditional Access/compliance, assign configuration/update ring and validate peripherals/guest join. Evidence: room acceptance checklist and sign-in evidence.

Lab 8 — Device and call-quality incident¶

Investigate one controlled sign-in/calendar/peripheral issue and one poor-media sample using Pro portal/TAC, Call Analytics, CQD, network and client/device evidence. Evidence: timeline, layer isolation, fix and retest.

Lab 9 — Shared Calling and emergency-location evidence¶

Tabletop user and resource-account routing separately, including inherited global usages, extension dialing, shared-number replacement, emergency callback ownership and the LIS example. If an authorized test environment exists, use only the carrier-approved test process. Evidence: user/resource policy matrix, correct client location, return-call route, negative cases and recovery ownership.

Lab 10 — Device-management and interop migration¶

Inventory app/OS/OEM/Agent/cloud versions, portal access and AOSP compliance. Reconcile settings-template imports and actual device state; test delegated scope, meeting join, required third-party capabilities and recovery after an unsuccessful configuration. Evidence: inventory, before/after settings, precise interop entitlement and feature tests; no completion claim based on devices merely appearing in PMP.

Knowledge checks¶

Teams Phone versus PSTN connectivity? Teams Phone supplies cloud calling control; Calling Plan, Operator Connect, Teams Phone Mobile or Direct Routing supplies the public-network path.
Why mix PSTN models? Different countries/personas may need different carriers, mobile integration, legacy connectivity or operating control.
Teams Phone Mobile best fit? Mobile-first users using one participating-operator SIM number across native mobile and Teams endpoints.
Shared Calling purpose? Eligible users place PSTN calls through a shared resource-account number without dedicated numbers.
Dial plan versus voice route? Normalize what was dialed versus select an outbound PSTN path.
PSTN usage purpose? Link a user's voice routing policy to eligible ordered routes.
SBC purpose? Securely interconnect Teams Direct Routing with carrier/PBX signaling and media.
SBA purpose? Preserve supported branch calling during cloud-connectivity loss.
LBR versus LMO? Regulatory/location routing constraints versus optimizing media through local SBC topology.
Why test Direct Routing failover? Redundant configuration does not prove DNS, TLS, routing, carrier and media recovery.
User versus service number? Person endpoint versus voice application/conferencing service under applicable model.
Resource account versus auto attendant? Identity/number anchor versus menu/schedule routing application.
Auto attendant versus queue? Menu/time routing versus distribution to agents.
Why test holiday/overflow? The normal business-hours path does not prove exceptional routing.
Verified emergency address enough? No; validate dynamic client location, route, notification and callback under approved procedures.
Meeting setting versus policy? Tenant-wide capability/default versus scoped organizer/participant control.
Template versus template policy? Reusable meeting configuration versus which users can select it.
Meeting versus webinar versus town hall? Collaboration, registration-led structured event and produced one-to-many broadcast.
Audio Conferencing versus Teams Phone? Meeting PSTN access versus ordinary user/application telephony.
Why resend an invitation after bridge changes? Existing invitation details may retain old conferencing information.
What affects meeting Copilot? Entitlement, policy/options, transcript/recording context, client and rollout.
Why eCDN? Reduce duplicate corporate-network streaming traffic for applicable large events.
QoS purpose? Prioritize media during contention; it does not create capacity.
CQD versus Call Analytics? Aggregate site/subnet/build patterns versus a user/session investigation.
Rooms Windows versus Android? Different compute/enrollment/management and sometimes feature timing; validate requirements.
Room resource account purpose? Device sign-in plus Exchange booking identity for a shared room.
Why not ordinary interactive MFA on Rooms? The shared appliance sign-in cannot satisfy unsupported user prompts; use supported CA/compliance controls.
Teams Rooms Pro management value? Eligible proactive room monitoring/operations beyond ordinary device administration.
Why update rings for rooms? Validate hardware, peripherals, sign-in and meeting behavior before fleet rollout.
SIP Gateway purpose? Connect supported SIP devices to Teams with documented feature limits.
BYOD room still needs governance? Yes; physical space, peripherals, discovery, booking and metadata remain managed.
Calendar missing but device online—first boundary? Inspect Exchange room mailbox/calendar processing separately from Teams sign-in.
One-way Direct Routing audio suggests? Media/NAT/firewall/SBC negotiation before dial-plan changes.
What proves a voice-user deployment? Effective objects plus inbound/outbound/features/emergency-approved validation and telemetry.
What proves a room deployment? Booking, sign-in, meeting join, content, every peripheral, guest join, update and monitoring acceptance tests.
Safest first incident action? Capture user/device, UTC time, call/meeting ID, number, policy, location, logs and recent changes before resetting.

Does Shared Calling remove the user’s Teams Phone requirement? No; it shares the number, not the user’s entitlement.
Which ordinary routing policy should be empty of usages? The Shared Calling user’s; a Direct Routing resource account needs the proper usages.
What is the SMS example’s excess? 150 measured segments over the pooled 2,000.
Does the most specific number pattern always win? No; PSTN usage order and route priority matter.
Is an SBC list an ordered failover pair? No; gateways within a route are tried in randomized order.
Is a valid address proof of dynamic location? No; inspect actual LIS match, client and propagation.
Can a site emergency policy override a direct user policy? Yes; ordinary Teams policy precedence cannot be generalized to this case.
Does a blind-transfer BYE prove the target answered? No; acceptance and target completion are distinct.
Does the AOSP profile QR code onboard a Teams room? No; Teams uses its account-based enrollment path.
Does deleting the processed XML prove every requested setting applied? No; compare effective supported settings, including possible partial application.
Do six dock events mean six distinct auto-association signals? No; the example contains only three users.
Does CVI enable every Teams Room to join every third-party meeting? No; direction, endpoint, plan, policy and platform feature support differ.

Final readiness checklist¶

I can choose a meeting/event, PSTN model, device and network architecture from requirements and explain its support boundary.
I can map every voice user, room and voice application to identity, license, number, policy, network and evidence.
I can build and troubleshoot dial plans, voice routing, auto attendants, call queues, emergency calling and Direct Routing.
I can configure policies/templates/settings for meetings, Audio Conferencing, webinars and town halls and prove participant behavior.
I can deploy and secure Windows/Android Rooms, phones, SIP devices, BYOD spaces and bookable desks through supported enrollment and access controls.
I can use Teams admin center, Rooms Pro Management, PowerShell, Call Analytics, CQD, device and SBC/carrier evidence without erasing the incident first.
I have completed the current blueprint checklist, hands-on labs and at least two timed readiness attempts using original, source-explained questions.

Places to learn¶

This is a curated starting point, not a complete list, and it is not meant to be consumed in full. Choose one primary route, build a lab/topology, and add resources only for measured gaps. Reconcile every older source with the retained April 28, 2026 outline and separately compare the future October 29 outline, especially Shared Calling, SMS, Queues app, Copilot, current events, Android enrollment/MDEP, flexible workspaces and room/device features.

October 28 learning route and evidence task¶

Start with the five-day Microsoft course and its two selected paths (15 modules, 7/8); choose voice or room investigations from actual weak areas. As of this review on October 28, 2026, the October 29 outline is still a future exam baseline. A course update or a five-day duration does not establish exam-transition timing. The existing Tom Arbuthnot/Marius Nilsen interop briefing is useful specialist context: its 2 min page label describes the synopsis, while video chapters extend past 15 minutes; the full video runtime was not verified or watched.

Original worksheet (about 30–45 minutes, editorial estimate): Draw an external SIP endpoint joining your Teams meeting and a Teams Room joining an external meeting. For each direction record endpoint/platform, provider/plan, join method, lobby expectation, required content/peripheral behavior, owner and one failure test. Mark any May 2026 rollout statement for a current support check.

Answer checkpoint: Similar meeting names do not establish the same direction, entitlement or feature set. An advertised interop capability needs a direction-specific acceptance test; reading the synopsis is not successful interoperability execution.

The two official paths are plan and design Teams collaboration communications systems (6h38) and manage Teams collaboration communications systems (8h05), a historical total of 14 hours 43 minutes. Current public pages list 15 modules (7/8) without those runtimes in retrieved text; use the previous timings only as planning estimates.

Resource	Access	Estimated time
Official MS-721 blueprint and credential page	Public	1–2 hours initially; 15 minutes per recheck
Two official paths from MS-721T00	Public	Historical 14h43; current 15 modules; allow 30–50 hours with labs/notes
MS-721T00 instructor-led course	Paid/partner delivery	5 days listed
Microsoft MS-721 Practice Assessment	Public	45–75 minutes per attempt plus source review
Pluralsight MS-721 path	Paid	Five public course listings total 8h04 (path rounds to 8h); 2023–March 2024 dates; paid lessons/practice unreviewed
O'Reilly/Apress MS-721 Certification Companion by Fabrizio Volpe	Paid	Historical 6h08 / 333 pages, October 2024; page blocked on recheck, current edition/content unverified
MeasureUp MS-721	Paid	167 questions (47/26/59/35), January 2025 update verified publicly; paid content/April completeness unreviewed; allow 6–10 hours
Microsoft Mechanics, Microsoft Reactor, and John Savill	Public	2–10 hours selectively; no complete current MS-721 playlist was confirmed
Partner Skilling Hub	Partner-restricted	Schedule dependent; use published start/end times after sign-in

Practice Assessment and video/partner sites returned public shells; no interactive questions or videos were reviewed. No exact current Whizlabs MS-721 offering was independently verified. Udemy listed several MS-721 question-only products, but this review did not establish enough provenance and freshness to recommend one. Start with Microsoft's free Practice Assessment; add paid questions only for a different explanation style or measured gap, and reject any provider claiming recalled live exam questions.

Blog readings with an output¶

Reading	Useful task and boundary
Teams Rooms interop comparison — Tom Arbuthnot with Pexip’s Marius Nilsen, May 11, 2026	Draw both calling directions and choose DGJ, CVI/SIP Guest Join or cross-platform SIP from requirements. The public synopsis is AI-assisted and team-reviewed; the video was not watched. Provider claims and older Android availability wording must be checked against current Microsoft documentation.
What’s new in Teams, July 2026 — Kerry Perez Heffernan, July 31	Turn Windows passwordless Rooms and Android settings-management announcements into a device/version/license/policy/recovery acceptance sheet. The main article was read; demos and actual device deployment were not tested.

MS-721 Collaboration Communications Systems Engineer Study Guide¶

Start here

How to use this guide¶

Objective map¶

1. Plan and design collaboration communications systems¶

Meeting and event architecture¶

Teams Phone and PSTN connectivity¶

Direct Routing, extensions and survivability¶

Choose a certified contact-center integration¶

Teams Rooms and certified-device design¶

Network and media design¶

2. Configure and manage Teams meetings, webinars, and town halls¶

Meeting policy stack and advanced experiences¶

Create and assign policy packages¶

Audio Conferencing¶

Webinars and town halls¶

Town halls across Teams, Viva Engage and recorded video¶

Configure avatars with an expiry boundary¶

Evaluate Teams Premium in the admin center¶

3. Implement and configure Teams Phone¶

User calling configuration and policies¶

Worked phone-policy matrix¶

Worked calling features — recipient and timing¶

Assign phone numbers in a controlled PowerShell batch¶

Auto attendants and call queues¶

Configure queue outbound identity¶

Dynamic emergency calling¶

Direct Routing implementation and troubleshooting¶

Configure on-network Audio Conferencing¶

4. Configure and manage Teams Rooms and devices¶

Resource accounts, enrollment and secure access¶

Operations and lifecycle¶

Configure Windows room display branding¶

Configure Android and panel display branding¶

Keep workstation GPOs off room appliances¶

Provision a supported SIP handset¶

Create and manage Teams device tags¶

Enable and verify hotline on a shared phone¶

Room features and flexible workspaces¶

Worked optional-room acceptance matrix¶

Create and publish room types and standards¶

Device troubleshooting¶

Integrated scenarios¶

Scenario 1: Multinational Teams Phone migration¶

Scenario 2: Headquarters Rooms sign-in failure after security rollout¶

Scenario 3: Executive town hall with remote sites¶

Hands-on labs¶

Lab 1 — Architecture and licensing matrix¶

Lab 2 — Network, CQD and eCDN plan¶

Lab 3 — Meetings, webinar and town hall controls¶

Lab 4 — Audio Conferencing and voice-user build¶

Lab 5 — Auto attendant and queue¶

Lab 6 — Direct Routing tabletop or lab¶

Lab 7 — Teams Room deployment¶

Lab 8 — Device and call-quality incident¶

Lab 9 — Shared Calling and emergency-location evidence¶

Lab 10 — Device-management and interop migration¶

Knowledge checks¶

Final readiness checklist¶

Places to learn¶

October 28 learning route and evidence task¶

Blog readings with an output¶