InsightPilot

Inspiration

In many teams, meetings generate a lot of discussion but very little execution. Action items are often mentioned casually, owners are left implicit, and deadlines are forgotten. After the meeting ends, context is lost and accountability weakens.

InsightPilot was inspired by this execution gap — the moment where good discussions fail to translate into completed work.

What it does

InsightPilot is an AI-powered meeting copilot embedded directly inside Jira issues. It continuously analyzes meeting discussions and issue comments to:

• Extract action items from conversations
• Detect missing owners and deadlines
• Surface unresolved follow-ups using contextual memory
• Provide clear recommendations to improve execution

Rather than acting as a passive note-taker, InsightPilot focuses on turning conversations into accountable tasks.

How I built it

InsightPilot is built as a native Jira app using Atlassian Forge. The app runs as an Issue Panel inside Jira, where it can access real issue context and comments.

The backend uses a resolver to fetch Jira issue data securely and analyze discussion text using lightweight heuristic intelligence. The frontend is built with Forge UI Kit, allowing the app to run seamlessly inside Jira without external infrastructure.

This approach prioritizes reliability, explainability, and real integration over simulated demos.

Challenges I ran into

The main challenges involved understanding Forge’s security model, handling Jira API access correctly, and ensuring the app followed platform constraints such as route-based API calls.

Another challenge was balancing intelligence with stability — avoiding over-engineering while still delivering meaningful insights within a hackathon timeframe.

What I learned

This project reinforced the importance of building real, deployable integrations rather than mock prototypes. It also highlighted how small, focused intelligence can deliver strong value when embedded directly into existing workflows.

What’s next

Future versions of InsightPilot will extend into real-time, in-call AI participation with continuous memory across meetings, proactive clarification during discussions, and deeper AI reasoning using large language models.

Built With

• forge
• javascript
• jiracloud
• node.js
• restapi