Inspiration
In agile teams, daily standups and project tracking are often repetitive, time-consuming, and prone to being overlooked. We wanted to automate and enhance the Scrum Master’s role using AI to make team communication smoother, project updates consistent, and task tracking effortless. Our inspiration was to build an AI-powered agent that doesn't just listen—but understands, responds, and keeps projects moving forward autonomously.
What it does
Scrum Maestro is an intelligent Slack-integrated Scrum Master bot powered by the uAgents framework and Gemini API. It automates key responsibilities of a Scrum Master:
Sends standup reminders to team members.
Parses natural language standup updates and extracts structured task information.
Tracks and updates task status (STARTED, IN PROGRESS, ENDED, ABANDONED).
Pushes tasks to MongoDB and creates linked Jira issues.
Generates daily standup summaries in a human-like tone.
DMs users about upcoming deadlines or missing updates.
Responds intelligently to queries and mentions in Slack with task-aware context.
How we built it
Tech Stack: Python, Slack SDK, MongoDB, Jira API, Google Gemini (via HTTPX), uAgents, dotenv.
Backend Agent: Built using the uAgents framework to handle events like periodic summaries, Slack polling, and deadline reminders.
Slack Integration: Utilizes Slack API to fetch messages, post reminders, and DM users. Uses message timestamps to prevent duplication.
NLP with Gemini: Google Gemini handles task extraction from natural language and summarizes team updates.
MongoDB: Stores structured task data along with user, status, and deadline.
Jira Integration: Automatically creates a Jira issue for every new task using Jira's REST API.
Challenges we ran into
Slack API Errors: Encountered not_in_channel errors which were resolved by ensuring the bot was invited to the Slack channel.
Parsing Natural Language Reliably: Gemini sometimes returned responses that weren't valid JSON, requiring robust regex and error handling.
Avoiding Duplicate Summaries: Ensuring timestamps were managed to not resummarize the same messages multiple times.
Time Zone Handling: Coordinating UTC-based timestamps for due dates and reminders across platforms.
Accomplishments that we're proud of
Built a fully autonomous Scrum Master agent capable of operating in a real team setting.
Successfully integrated multiple APIs (Slack, Jira, Gemini, MongoDB) into one cohesive agent.
Made daily standups asynchronous, structured, and insightful without manual effort.
Developed a smart, contextual AI that feels human and helpful—not robotic or redundant.
What we learned
How to build production-ready multi-agent systems using the uAgents framework.
Effective Slack bot design and message polling.
Real-world NLP application using large language models for structured task extraction.
Jira API integration and how to automate project management workflows.
How to build trust and context into AI responses so teams feel supported, not spammed.
What's next for Scrum Maestro
UI Dashboard: A web dashboard to visualize team progress, burn-down charts, and task timelines.
Voice Integration: Integrate with tools like Google Meet or Zoom to collect standups via voice.
Intelligent Task Delegation: Use organizational role data to suggest and assign tasks to the right people.
Sprint Analytics: Add automatic generation of sprint summaries and progress analytics.
Customizable Personas: Allow teams to customize the tone and behavior of their ScrumMasterBot.
Log in or sign up for Devpost to join the conversation.