JiraGenie

Inspiration

Navigating Jira with filters and queries can slow teams down. We wanted a faster, more natural way to interact with issues by simply talking to Jira and getting instant, AI-driven insights.

What it does

JiraGenie lets users ask questions about Jira issues using text or voice. It understands natural language and provides AI-generated summaries, and insights all directly inside Jira.

How we built it

We built JiraGenie on Atlassian Forge using the Forge UI Kit for the frontend and Google Gemini for AI-powered understanding. ElevenLabs handles real-time speech-to-text transcription for voice input. The backend uses Forge resolvers to securely fetch Jira issue data and pass it to Gemini for analysis.

Challenges we ran into

Integrating ElevenLabs’ real-time transcription with Forge’s limited UI event model
Rendering dynamic AI responses within the constraints of Forge UI Kit
Managing secure communication between Jira, Gemini, and ElevenLabs APIs

Accomplishments that we're proud of

Created a fully voice-enabled Jira assistant inside Forge
Seamless AI integration with Gemini for contextual analysis
Reliable transcription pipeline using ElevenLabs’ streaming API

What we learned

We learned how to push the boundaries of Forge UI Kit with external API integrations and real-time updates. We also gained experience building multimodal (voice + text) AI tools with strong security practices.