AI project management bot case study · Microsoft Teams · Azure DevOps · IT delivery

A bot that turns Teams meetings into reviewed DevOps backlogs

This AI project management bot case study covers an IT and software delivery organisation that runs sprint-based projects on Microsoft Teams and Azure DevOps. Meeting notes, task allocation and sprint tracking were manual, which caused delays and inconsistencies. We built a bot that joins Teams meetings, an AI agent that extracts tasks from the transcript, and a manager review step before anything reaches Azure DevOps. The MVP is delivered and Phase 1 is implemented.

Agent run · Sprint planning meetingRunning
Join
Transcript
Extract
Review
Backlog
Draft backlog from one sprint meeting
Source
Teams meeting transcript
Draft
Epics, User Stories, Tasks
Sizing
Fibonacci story points
Target
Nearest sprint
M
Task review before creation in Azure DevOpsManager
ApproveHold
Nothing posts until a person approvesIllustrative
Before and after

What changed for the delivery team.

BeforeWith the agent
Meeting notesCaptured by hand, which caused delays and inconsistenciesA bot joins the Teams meeting and captures the recording and transcript
Task extractionSomeone read the notes and wrote up the tasksAn AI agent pulls objectives and tasks from the transcript
Backlog creationEpics, stories and tasks entered manuallyDraft Epics, User Stories and Tasks are generated for Azure DevOps
Task assignmentAllocated by handTasks go to the right team and developers, using story points and sprint capacity
Sprint visibilityManagers lacked real-time visibility into velocity and pending tasksPower BI dashboards show velocity, progress trends and pending tasks
ControlManager did everythingManager reviews, edits or deletes tasks before approval
How it works

From Teams meeting to approved backlog.

The agent drafts. The manager decides. Nothing lands in Azure DevOps until the manager approves it.

1 · What comes in
Microsoft Teams meeting
Recording and transcript
RAG data for priorities
2 · AI project management agent
Bot joins the meeting through a dedicated service account
Recordings and transcripts captured with Power Automate and the Fireflies.ai API
Azure AI and Azure AI Search read the transcript and extract objectives and tasks
RAG data sets priorities; vector indexes support retrieval and search
Tasks mapped to Backend, Frontend, QA, Designer or Project Manager
Fibonacci story points and sprint capacity guide assignment
3 · Manager Review Portal
Draft Epic with its User Stories
User Story with Tasks for the Backend team
Task for QA waiting for review
Task the manager edited or deleted
Manager
4 · Posted in
Microsoft Teams
Power Automate
Azure DevOps
Power BI
Microsoft Graph API
Generated tasks are emailed to the manager with a portal link, and approved tasks are created in Azure DevOps automatically.
Results

What the MVP delivers today.

Automated backlog creation

Epics, User Stories and Tasks are drafted from the meeting transcript and placed in the nearest sprint once approved. Only tasks from that meeting are added.

Fair workload distribution

Tasks are matched to the right team and assigned using Fibonacci story points and available sprint capacity.

Real-time sprint tracking

Power BI dashboards cover velocity, progress trends and pending tasks, and email reminders go out for incomplete tasks.

Why it worked

A person stays in control.

Manager approves first

Tasks can be reviewed, edited or deleted in the Manager Review Portal. Azure DevOps only gets approved work.

Secure access to Teams data

An Azure AD app with an Application Access Policy and the Microsoft Graph API control access to Teams data.

Built on the tools they already use

Teams, Azure DevOps, Azure AI and Power BI are connected, so the team keeps its existing way of working.

FAQ

About this project.

Is this a finished rollout?

No. The status is MVP completed, with Phase 1 implemented.

Does the agent create tasks on its own?

No. Generated tasks are emailed to the manager with a portal link. The manager reviews, edits or deletes them, and approved tasks are then created in Azure DevOps.

How are meetings captured?

A bot joins the Teams meeting through a dedicated service account. Recordings and transcripts are captured with Power Automate and the Fireflies.ai API and stored in a SQL database.

How are tasks assigned?

Tasks are matched to the right team (Backend, Frontend, QA, Designer, Project Manager). Developers are then chosen using Fibonacci story points and available sprint capacity. The sprint rule, Sprint 1 as day 1-15 and Sprint 2 as day 16 to month end, is specific to this setup.

Sprint admin eating your week?

Want an agent that drafts your backlog from meetings?

Tell us how your team plans sprints today. We'll show you where an agent can draft the work and your manager still approves it.

Book Appointment
Sahil kataria (1)
Sahil Kataria

Founder and CEO

amit Kumar
Amit Kumar

Chief Sales Officer

Talk To Sales

USA

+1 270-550-1166

flag

+1 270-550-1166

Phil J.
Phil J.Head of Engineering & Technology​
QServices Inc. undertakes every project with a high degree of professionalism. Their communication style is unmatched and they are always available to resolve issues or just discuss the project.​
Assured

Thank You

Thanks, we’ve got your details. A named member of our team will reply within one business day.