Agile Analytics — Sprint Intelligence for Azure DevOps

15+ analytics views built directly into Azure DevOps — Flow Efficiency, cycle time with auto-discovered custom states, Service Level Expectations on the dashboard, Monte Carlo forecasting, WIP monitoring, sprint capacity, and AI adoption tracking. No infrastructure. No spreadsheets. No BI tools.

Install in 2 minutes. See health scores, delivery forecasts, and flow data for every team in your organization — live, inside ADO, from your first sprint.

At a glance

What it is. A sprint-intelligence and flow-analytics hub embedded in Azure DevOps Services. 15+ analytics views and 3 native ADO dashboard widgets, delivered as a single project hub extension.
Built for. Engineering managers, agile coaches, scrum masters, and product owners running 1–50+ Scrum, Kanban, or SAFe teams in Azure DevOps Services. Useful from a single team to enterprise rollouts.
Pricing. Flat org-wide subscription. $1,500 / year Founding Partner (limited time) · $2,000 / year Annual · $200 / month. Unlimited users, teams, and projects in your organization. 30-day free trial activates automatically on install — no credit card.
Security & privacy. All analytics run in your browser against the Azure DevOps REST API using your existing ADO session. Your work-item data never leaves your Azure DevOps tenant. Baytek's backend stores only licence records and lightweight install telemetry — never work-item content.
Compliance. WCAG 2.1 AA accessibility achieved (verified by an axe-core regression suite in CI on every release). SOC 2 Type 1 evidence collection in progress. GDPR-aligned data handling — see Privacy Policy.
Permissions. Three read-only Azure DevOps scopes: vso.project, vso.work, vso.graph. No vso.code, no vso.analytics, no write scopes. The extension cannot modify your work items.
Quality bar. Every release passes 500+ automated tests (vitest unit + Playwright E2E + 5-view accessibility regression baseline + visual regression baseline) before publication. TypeScript strict mode end-to-end.
Infrastructure required. None. No Azure Function, no database, no service connection, no PAT. Install, map workflow states once, done.
Where it runs. Azure DevOps Services (cloud). Azure DevOps Server (on-premises) is not supported in this version.
Publisher. Baytek. Support and feature requests: ado-analytics.baytekdev.com/support

The Gap It Fills

Azure DevOps has powerful raw data but no built-in answers for the questions that actually matter in sprint planning and retros:

"When will we be done?" — Monte Carlo simulation on your actual throughput history
"How much can we realistically commit this sprint?" — Sprint Capacity: velocity-adjusted commitment, factoring in each member's days off
"Why is this sprint struggling?" — Commitment reliability, scope churn, and carryover rates in one view
"Which team needs attention?" — Cross-team health grid with configurable goals and auto-generated coaching prompts
"How long does work actually take?" — Cycle time scatterplot with P50/P85 percentile lines
"Are we flowing or are we stuck?" — Flow efficiency, lead time, and WIP trends over rolling windows
"Are our engineers actually using AI tools?" — Company-wide GitHub Copilot and Microsoft 365 Copilot adoption dashboard

Agile Analytics answers all of these — without leaving Azure DevOps.

Who Is This For

Role	What you get
Scrum Master	Live sprint health, commitment reliability, scope churn alerts, sprint capacity planning, and AI-generated coaching questions per team
Engineering Manager	Multi-team velocity comparison, cycle time trends, flow efficiency, team leaderboard, and AI adoption metrics
Agile Coach	Configurable health thresholds, prioritized action items (P1/P2/P3), and a retro snapshot you can export to PowerPoint in one click
Product Owner	Feature delivery progress, Epic/Feature completion with child story rollups, and Monte Carlo forecast dates
Developer	Personal metrics: PRs merged, review participation, work items closed — across any sprint range

Views at a Glance

Dashboard — Redesigned hub with sprint KPI cards, burndown, WIP, and delivery grade at a glance

Sprint Health

View	What it shows
Dashboard	Sprint KPI summary: velocity, cycle time, burndown status, WIP count, and delivery grade
Sprint Capacity	Per-member availability based on ADO capacity settings, team days off, effective capacity %, and velocity-adjusted recommended commitment
Live Stats	Real-time item counts by workflow state for the active sprint — refreshes on every open
Sprint Summary	Single-page recap of velocity, completion %, scope churn, carryover, and delivery grade — ready to drop into a retro
Multi-Team Scope	Side-by-side completion % across all your teams in one scrollable view
Delivery Signals	Commitment reliability %, scope churn %, and carryover rate per team with trend direction

Live Stats — Real-time workflow state counts for the active sprint

Sprint Summary — One-page recap with velocity, scope churn, carryover, and delivery grade

Flow Analytics

WIP Monitor — Active items with age and configurable team WIP limits

View	What it shows
WIP Monitor	Active items with age in days, configurable team WIP limits, and breach alerts with Teams/Slack notifications
Flow Metrics	Throughput, lead time, and flow efficiency trends over rolling 4/8/12-week windows
Flow Efficiency	Touch time vs. wait time per state — see exactly where your cycle time is going. Top-25 items table sorted by cycle time. Healthy knowledge work runs 15–40%.
Aging & Blocked	Items exceeding configurable age thresholds, with blocked items flagged separately
Cycle Time	Interactive scatterplot — completion date vs. days-to-complete, with P50/P85 percentile lines
Monte Carlo	Run 10,000 simulations on historical throughput — get P50/P85/P95 delivery date estimates

Workflow Mapping — Auto-discovered project states with a live cycle-time preview panel

Portfolio

View	What it shows
Team Leaderboard	Normalized ranking across all teams: velocity, reliability, scope discipline
Feature Analytics	Epic and Feature cards with live child-story completion bars and rollup counts

Monte Carlo Forecasting

Monte Carlo — 10,000-run simulation returning P50/P85/P95 delivery date estimates

Agile Coach

View	What it shows
Agile Report	Multi-team health grid with configurable goals, status badges (On Track / At Risk / Critical), and auto-generated P1/P2/P3 coaching action items
Multi-Team Aging	Cross-team view of items exceeding age thresholds, by team and state
User Metrics	Per-contributor view: PRs merged, code reviews given, work items closed — filterable by sprint range

AI Governance

View

What it shows

AI Metrics

Company-wide AI tool adoption dashboard. Stat cards: Active Users, Acceptance Rate, Suggestions Generated, Teams Active. Tabs: Overview (trend chart, Setup Roadmap, Custom KPIs), Users (top/bottom 10 by acceptance rate), Teams (team leaderboard), Insights (actionable adoption insights), Settings (connect GitHub Copilot, Microsoft 365 Copilot, or a Custom API endpoint). Shows realistic demo data until a live source is connected.

AIIP

AI Improvement & Intervention Panel — admin-only. Severity-coded alerts, intervention tracking with official Microsoft Learn links, milestone celebrations, per-team Teams webhook notifications, and a full audit log.

Dashboard Widgets

Three Azure DevOps dashboard widgets — add them to any project dashboard alongside your other ADO widgets:

Widget	Sizes	What it shows
Team Health	1×2, 2×2	Health score, grade, and throughput trend for a selected team
Cycle Time	2×2, 2×4	P50 and P85 cycle time trends over the last 6 sprints
Throughput	2×2, 2×4	Throughput per sprint bar chart for the last 6 sprints

Tools

View	What it shows
Retro Snapshot	One-click PowerPoint export of the current sprint summary, ready for your retro
AI Assistant (optional)	Ask natural language questions about your sprint data — connect your own Claude or OpenAI key
Configuration	WIP limits, workflow state mappings (auto-discovered), Service Level Expectations, sprint alerts, notification webhooks, background monitors, access control, readiness checks, support diagnostics

Dark mode — Full dark theme, one-click toggle, persisted per user

How It Compares

Capability	Agile Analytics	Native ADO Analytics	Power BI dashboards	Typical Jira marketplace plugin
Cycle time scatterplot (P50 / P85 percentile lines)	✓ built-in	—	requires data modeling	varies; often paid extra
Monte Carlo forecasting (10,000-run simulation)	✓ built-in	—	external add-on	extra licence
Flow Efficiency (per-state touch vs. wait time)	✓ per-state breakdown	—	manual model	rarely available
Sprint Capacity from ADO Boards capacity	✓ velocity-adjusted commitment	partial	manual	n/a
Multi-team health grid + AI-generated coaching prompts	✓	—	manual	n/a
WIP alerts to Microsoft Teams / Slack	✓ in-product + Background Monitor pipeline	—	external	varies
Setup time	<2 minutes	included	days to weeks	varies
Pricing model	Flat org subscription	included with ADO	per-user Power BI licence	per-user
Infrastructure required	None	None	Power BI workspace + dataflows	Plugin host
Work-item data leaves your tenant?	No — browser-direct to ADO REST	n/a	Yes (ingested into Power BI)	Often yes

This is positioning, not a benchmark — every team's setup is different. The point is that Agile Analytics fills the analytics layer that ADO ships without, with no data-pipeline work and a flat org price.

Security, Privacy & Compliance

Where your work-item data lives. Every analytics view runs in your browser against the Azure DevOps REST API, authenticated by your existing ADO session — the same session ADO Boards uses to render your work-item list. Baytek does not see, request, or store your work-item content, sprint data, AI prompts, or any user identifiers from your tenant.

What Baytek's backend handles.

Licensing. When an org admin activates a paid licence, the backend issues a signed activation token. Subsequent validation calls confirm the token is still valid. Stored: organisation name, plan tier, expiry date, activation token. No work-item data.
Install telemetry. A lightweight ping on first install and on subsequent loads (organisation name, ADO org id, extension version, event type, timestamp) lets us count active orgs and detect outages or regressions. No work-item data.
Optional trial-end reminder. If a trial admin opts in by entering their email in the trial reminder card, Baytek stores that email solely to send one trial-ending notification. Opt-in only.

Sub-processors.

Stripe — payment processing for licence purchases.
Firebase Hosting — the ado-analytics.baytekdev.com marketing and licence-management site only. The extension itself does not depend on Firebase.
Zoho Mail — transactional email (licence keys, support replies).

Sub-processors are listed in the Privacy Policy and the compliance/ working evidence record.

Compliance posture.

Accessibility. WCAG 2.1 AA achieved as of v6.2.4. Every release runs a 5-baseline axe-core regression suite in CI; new violations fail the gate before publication. The single accepted exception is page-has-heading-one, which is structurally inappropriate for an iframe-embedded extension whose host page (Azure DevOps) owns the page-level h1 — adding our own would create competing top-level headings in the user's accessible heading tree.
SOC 2 Type 1. Evidence collection in progress. Working documentation maintained internally; ask via support for status updates.
GDPR. Personal data processed by Baytek is limited to the licence record and install telemetry above. Customers acting as data controllers can request deletion via support.

Settings storage. All extension configuration (workflow mapping, WIP limits, AI provider keys, access control, user preferences) is stored at organisation scope in Microsoft's Azure DevOps Extension Data Service — Microsoft-managed infrastructure that Baytek systems do not access. Configuration survives extension upgrades and is removed by Microsoft on uninstall.

Engineering quality bar. Releases are gated through 500+ automated tests:

Unit tests (vitest) — exercise every analytics calculation against real-ADO captures and synthetic fixtures across all four Microsoft stock processes (Agile, Scrum, CMMI, Basic) plus arbitrary custom processes.
End-to-end tests (Playwright) — validate the hub shell, navigation, configuration save persistence, trial flows, accessibility, and visual regression baselines.
Accessibility regression suite — axe-core runs against five baseline views on every release; new WCAG 2.1 AA violations fail the gate.
Visual regression baselines — five baseline screenshots gated for unintended UI drift.

A failure in any layer stops the release from being packaged.

What's New in Version 6.2

v6.2.5 — Security update (May 2026)

Server-authoritative trial enforcement, tightened licence-validation cache (cached "licensed" state expires after 7 days without a successful re-validation), and a security release-notes pop-up. Existing licences and trials are preserved; no customer action required.

v6.2.4 — Accessibility + persistence

Complete WCAG 2.1 AA color-contrast pass: active nav pill, theme toggle, NPS feedback button, amber callout boxes, Cycle Time bucket pills, Configuration sidebar, Configuration active tab, and disabled-state context selector. Project / team selection now persists across reloads — your saved preference wins over the URL context.

v6.2.3 — Per-ADO-type sub-pills on Cycle Time

If you map multiple ADO types into the same analytics bucket (e.g. Bug + Issue both into Bugs), each type now gets its own sub-pill on the Cycle Time chart so you can isolate one at a time and see its own P50 / P70 / P85 / P95.

v6.2.2 — Item Type Mapping (3-bucket model)

Three fixed analytics buckets — Planned Work, Bugs, Exploration — drive every chart. Item Type Mapping is the customer-facing config that funnels every ADO work-item type into one of those three. Unmapped types now surface as an orphan warning instead of being silently dropped from charts.

v6.2 — Auto-detected mappings + "what's counted" banners

New installs scan your ADO project and pre-fill suggested item type mappings — review on the Item Types tab and save. Each chart now shows exactly which ADO type names feed each category, with a yellow warning when items are excluded.

Read the full changelog →

What's New in Version 6.1.9

Polish for the redesigned hub plus opt-in trial reminders.

Refreshed Dashboard, Live Stats, Sprint Summary, WIP Monitor, and Workflow Mapping The new screenshots above are the actual UI you get on install — same data and metrics as before, restyled with the Primer-aligned design system. Live Stats now refreshes the moment you open it, the Sprint Summary is one scroll-free page, and the Workflow Mapping screen shows your auto-discovered ADO states with the cycle-time preview panel inline.

Trial-end reminder (opt-in) Trial users in the last 14 days, and orgs whose trial has just expired, see a small in-product card to leave an email address. We'll send one reminder if the trial is about to end — no marketing, and the licence still has to be activated by an Azure DevOps organisation admin. The card is dismissible and snoozes for 7 days.

Dark-mode polish across licence flows Trial banner, licence-locked screen, and the Configuration → License tab now render correctly in dark mode.

Bug fixes & small improvements General fixes to licence validation timing, mock SDK behaviour for local dev, and tightened webpack output for faster cold loads.

Read the full release notes →

What's New in Version 6.1

Closes the loop between your Azure DevOps process and the metrics on your dashboard.

Custom states now auto-discovered Workflow Mapping reads your project's actual states from Azure DevOps. Any custom state your team uses ("Awaiting QA", "Stakeholder Sign-off", etc.) shows up in the dropdown automatically — no more typing them by hand.

Cycle Time preview inside Workflow Mapping A new green panel under each project mapping shows exactly which of your states count toward cycle time (active + queued) and which are excluded (pre-work like New / Approved). The mapping IS the cycle-time configuration — no separate place to look.

Flow Efficiency view Under Measure → Flow Efficiency. Shows the percentage of cycle time spent actively working vs. waiting in queues, broken down per state, with a top-25 items table sorted by cycle time. Healthy knowledge-work teams typically run 15–40%.

Service Level Expectations (SLEs) Configuration → SLEs. Set per-item-type cycle-time targets (e.g. "85% of User Stories finish in 14 days"). The Dashboard shows a green / amber / red badge for each enabled SLE so the team can see at a glance where flow is on track and where it isn't.

"How is this calculated?" tooltips A small (?) icon next to Cycle Time and Flow Efficiency headers explains the math and links straight to Workflow Mapping.

Readiness check covers Workflow Mapping Configuration → Readiness now flags projects without a workflow mapping configured, so cycle time / lead time / Flow Efficiency aren't silently zero.

Read the full release notes →

What's New in Version 6.0

Major workflow-mapping improvements + a redesigned interface — please re-check your mapping after upgrading.

Redesigned interface A new left-side navigation, project + team switcher in the sidebar, and a dashboard with live KPI cards, sprint burndown, WIP, and sprint health at a glance. Every view you used before is still here — find them grouped under Live, Health, Plan, Measure, Agile Coach, and Tools in the sidebar.

Workflow mapping: explicit N/A for unused stages Each canonical stage (New, Queued, In Progress, Test, Review, Done) can now either be mapped to one or more ADO states or explicitly marked "— N/A (this stage is not used) —" when your process doesn't track it. N/A stages no longer surface as "unmapped" warnings in dashboards or readiness checks.

Aging Chart now detects truly blocked items — three ways Items show up in the Aging Chart's Blocked tab automatically when ANY of these match: the ADO state name contains "block" / "impediment" / "on hold", an ADO tag of "blocked" / "impediment" / "on hold", OR a bracketed [blocked] / [impediment] / [on hold] marker in the title. Stalled items (above your age threshold) still show too, with a clear chip telling you which signal triggered the entry.

Sprint Capacity opens the right sprint in ADO The "Open Sprint in ADO Boards" link in the empty state now deep-links straight to the correct sprint's Capacity tab for your selected team — no more wrong-org redirects.

Dark mode polish Configuration tabs, Org Hygiene panels, and Sprint Capacity now render correctly in dark mode (no more white panels in a dark frame).

After upgrading, please:

Open Configuration → Workflow Mapping and re-check your mapping per work item type — N/A is the new way to skip stages your process doesn't use.
If you previously worked around the lack of "blocked" detection by tagging titles, you can now move those items to a state with "blocked"/"impediment"/"on hold" in the name and they'll surface automatically.

Read the full release notes →

What's New in Version 5.0

Sprint Capacity Planning Know exactly how much your team can commit before you plan — not after you overcommit.

Reads capacity directly from Azure DevOps Boards (no duplicate data entry)
Per-member availability: personal days off + team-wide days off, de-duplicated
Effective capacity % with sprint working day count
Velocity-adjusted recommended commitment: your average throughput × effective capacity

Simplified Licensing — Flat Org Pricing One subscription covers your entire organization — unlimited users, teams, and projects. No per-seat counting, no ADO marketplace billing. 30-day free trial activates automatically on install, no credit card required.

In-Product Feedback Trial users see a lightweight feedback card — share what to build next or request features directly from within the extension.

Why Teams Choose Agile Analytics

No infrastructure — ever Everything runs in the browser against your live ADO data. No Azure Function, no database, no service connection to configure. Install, claim admin access, map your workflow states once, and it works across your entire organization.

Sprint Capacity + Monte Carlo together Plan with capacity-adjusted commitments, then validate with a 10,000-run Monte Carlo forecast. The two views are designed to be used together before every sprint.

WIP alerts that actually reach you Set WIP limits per team, connect a Teams or Slack webhook, and get notified the moment a team goes over — with a direct link to their active sprint taskboard.

Background Monitor — alerts without the browser open Configuration generates the Azure Pipeline YAML for you. Your admin adds it to a repo, sets a cron schedule, and WIP alerts run automatically — even when nobody has ADO open. Zero external infrastructure, runs entirely within your tenant.

Retro-ready in one click Export any sprint summary to a branded PowerPoint slide deck. Paste it into your retro without formatting work.

AI that uses your actual data When the AI Assistant is enabled, it reads your live sprint report before answering. Not generic advice — answers grounded in your team's actual numbers. Connect Claude, OpenAI, or GitHub Copilot using your own API key.

Role-based access control Built-in Admin/User role system with ADO directory search. Keep it open to all project users, or switch to assigned-users-only mode for controlled rollouts.

Dark mode Full dark theme — persisted per user, toggled with one click.

Getting Started in Under 2 Minutes

Step 1 — Install Open the Marketplace listing and install the extension into your Azure DevOps Services organization.

Step 2 — Open Navigate to any project → find Agile Analytics in the left navigation bar under your project name.

Step 3 — Claim Admin Access (Required) An Azure DevOps Project Administrator or Organization Administrator must go to Tools → Configuration → Access Control and click Claim Admin Access.

Step 4 — Configure Workflow Mapping (Required) Go to Tools → Configuration → Workflow Mapping and map your board states to analytics flow stages. This is what makes cycle time and flow metrics accurate for your process.

Step 5 — Select your team and explore Use the team selector in the top-left corner, then browse views using the top navigation.

Optional Visit Tools → Configuration to set WIP limits, add Teams/Slack webhook alerts, configure the Background Monitor, and run Readiness checks.

Complete Access Control claim and Workflow Mapping before broad rollout for correct analytics behavior.

Pricing & Licensing

One plan for your whole organisation: $2,000 / year or $200 / month. Unlimited users, teams, and projects. All features included. 30-day free trial — no credit card required.

Install from the Azure Marketplace and your trial starts automatically. When you are ready to subscribe, visit https://ado-analytics.baytekdev.com/pricing/. After payment, Baytek emails your licence key and an org admin activates it in Settings → License inside the extension.

Limited-time Founding Partner offer: $1,500/year locked in forever for the first customers before the offer closes. See the pricing page for details.

Support & Feature Requests: Questions, bugs, rollout help, or missing features? Start here: https://ado-analytics.baytekdev.com/support/

Access Control

Agile Analytics includes a lightweight built-in role system — no Azure DevOps group configuration required.

An Azure DevOps Project Administrator or Organization Administrator visits Configuration → Access Control and clicks Claim Admin Access
Admin chooses whether the extension is open to All project users or restricted to Assigned users only
Admin adds users by searching the ADO directory and assigning them Admin or User roles
Admin configures which views are visible to User-role members
Admins always see every view regardless of settings

Settings are stored at organization scope in Azure DevOps Extension Data Service — one configuration shared across all users in the org.

AI Assistant (Optional)

Disabled by default. When enabled:

Ask natural language questions: "How did Team Alpha perform this sprint?"
Reads your live sprint data before answering
Connect Anthropic Claude, OpenAI, or GitHub Copilot with your own organization-managed API key
Configure provider, model, per-session request limits, and output token limits
Admins can disable it org-wide via Configuration → Access Control

Data sent when AI is active: your prompt + sprint name + team name + aggregated metrics (point counts, percentages). Work item titles and user names are never included automatically.

Notification Alerts

Connect to Microsoft Teams or Slack to receive automatic alerts. Seven configurable triggers:

Sprint Health

Commitment Reliability Low — fires when reliability drops below your threshold
Scope Churn Exceeded — fires when mid-sprint scope changes exceed your churn threshold
Carryover Risk — fires when carry-over share is too high
Sprint End Summary — one-time digest at sprint end

Flow & WIP

WIP Over Limit — fires when any team exceeds its WIP limit; links directly to their active sprint taskboard
Stale In-Progress Items — fires when items have been in-progress for more than 5 days

Digest

Weekly Sprint Digest — once-per-week summary of sprint health and flow status

Configure in Tools → Configuration → Notifications. Paste your webhook URL, select triggers, and save.

In-session vs Background: Alerts above fire when a user has the extension open. For automated alerts without the browser open, use the Background Monitor.

Background Monitor — Alerts Without the Browser Open

Configuration → WIP Settings generates a YAML file for your Azure Pipeline. After your admin adds that file to a repo and creates the pipeline, it runs on your chosen cron schedule, checks WIP limits, and posts to Teams or Slack automatically.

Zero external infrastructure — runs entirely within your ADO organization
No PAT required — uses Azure Pipeline's built-in service account
Fully auditable — YAML file is committed to your repo; inspect or modify anytime

Permissions Explained

Permission	Why
`vso.project`	Read your project list and team roster for the team selector
`vso.work`	Read work items, sprint iterations, capacity, and work item history for all analytics views
`vso.graph`	Search the ADO user directory only when an admin adds a user in Access Control

We request only what we use. No vso.code, no vso.analytics, no write permissions.

Requirements

Azure DevOps Services (cloud) — any tier
Project Contributor access or higher
Modern browser (Edge, Chrome, Firefox)
Azure DevOps Server (on-premises) — not supported in this version

FAQ

Does this extension store my data anywhere? Your Azure DevOps work item content, sprint data, and AI prompts stay in your browser and your ADO tenant — Baytek never sees them. The only data on Baytek systems is your licence record (organisation name, plan, expiry, activation token) and a lightweight install/heartbeat ping (organisation name, extension version, timestamp) so we can count active installs and detect outages. Settings, workflow mappings, AI API keys, and access control are stored in your org's Azure DevOps Extension Data Service — Microsoft-managed infrastructure that Baytek does not access. Full details in the Privacy Policy.

Will my settings be lost when the extension updates? No. All settings (WIP limits, workflow mappings, notification webhooks, AI configuration, access control, user preferences) are stored using the ADO Extension Data Service key-value API, which is version-agnostic. Upgrading never resets your configuration.

Does it slow down Azure DevOps? No. The extension makes the same REST API calls you would make manually — one view at a time, only when you navigate to it. No background polling, no persistent connections.

Why do I see no data on Cycle Time, Flow Metrics, or Flow Efficiency? These views require completed sprints with items in a Done/Closed/Resolved state. As of v6.1.0 your project's custom state names are auto-discovered — open Configuration → Workflow Mapping and assign each of your states to a stage (In Progress / Test / Review / Done). The Cycle Time preview panel under each mapping shows exactly which states will count.

How does Sprint Capacity work? Sprint Capacity reads the capacity your team sets in Azure DevOps Boards (the days/hours per member per sprint). It calculates each member's available working days after personal and team-wide days off, computes effective team capacity as a percentage, and multiplies your average velocity by that percentage to suggest a realistic sprint commitment. If capacity hasn't been set in ADO Boards, the view will prompt you to add it there first.

Do I need to configure anything before rollout? Yes. An admin should claim admin access (Configuration → Access Control) and complete workflow mapping (Configuration → Workflow Mapping) before rolling out to the team.

How do I set up the AI Assistant? Go to Configuration → AI Settings. Select Claude, OpenAI, or GitHub Copilot, enter your API key, set usage guardrails, then enable and save.

Can I control who sees what? Yes — through the built-in Access Control. Admins can choose open access for all project users or restrict to assigned users only, and control which views regular users can see.

Does it work across multiple projects? Yes. The extension installs at organization level and is available in every project. The current plan includes unlimited teams and projects across your organization.

Is dark mode supported? Yes. Click the theme toggle in the top-right of the navigation bar. Preference is saved per user.

What happens if I uninstall? Your ADO work items, sprints, and boards are completely unaffected — this extension only reads data, never writes to work items. Extension Data (settings, preferences, access control) is deleted on uninstall.

What is AI Metrics and who can use it? AI Metrics is a company-wide AI adoption dashboard. It is included in the 30-day trial and every paid subscription. When no live AI source is connected, it shows realistic demo data so you can explore the layout. To connect a live source, go to AI Metrics → Settings (admin required) and enter credentials for GitHub Copilot, Microsoft 365 Copilot, or a Custom API endpoint.

What is AIIP? AIIP (AI Improvement & Intervention Panel) is an admin-only companion to AI Metrics. It surfaces automated alerts when users are at risk of disengagement (low usage, high rejection rate, sudden drop), lets admins log interventions and mark them done, celebrates milestones, and can send Teams webhook notifications per team. All actions are written to an audit log.

What are the dashboard widgets? Three Azure DevOps dashboard widgets: Team Health, Cycle Time, and Throughput. Add them to any project dashboard from the widget catalog — they appear under the "Agile Analytics" category. Each widget uses the same team selector and data as the main hub views.

Data & Privacy

What stays in your tenant: Azure DevOps work item content, sprint data, AI prompts and responses, and any AI provider API keys you enter. Baytek never sees any of this.

What Baytek's backend receives: licence records (organisation name, plan, expiry, activation tokens) and a lightweight install/heartbeat ping (organisation name, extension version, event type, timestamp). Plus an optional admin email if a trial admin chooses to opt in to a trial-end reminder.

Optional integrations: when you enable the AI Assistant or webhook notifications, requests go from your browser directly to the AI provider (Anthropic / OpenAI / GitHub Copilot) or webhook URL (Microsoft Teams / Slack) — never via Baytek.