Lolae

Inspiration

We kept losing context across Slack, Jira, and meetings. Standups repeated what tools already knew. We wanted one assistant that reads the noise and hands back a clear daily brief.

What it does

Lolae connects to Slack, Jira, and Google Meet, then generates AI briefings so teams start the day knowing what changed, what is blocked, and what needs them. It turns scattered updates into a single readable summary and surfaces relevant Jira tickets inline.

How we built it

A backend service ties the integrations together with a multitenant model (users, accounts, and generic connections). Slack is account scoped, Jira and Meet are user scoped. We use Gemini to summarize activity and format the briefings, and run everything on Google Cloud.

Challenges we ran into

Getting multitenancy right was the hardest part: mapping identities cleanly across three providers without leaking data between accounts. Formatting AI output into consistent, skimmable Slack and Jira views also took real iteration.

Accomplishments that we're proud of

A working end to end pipeline from raw tool activity to a polished daily brief, a clean multitenant connection model, and briefings that people actually read instead of skip.

What we learned

Integrations are easy to demo and hard to make trustworthy. Clear data boundaries and consistent formatting matter more than adding features.

What's next for Lolae

More connectors, smarter prioritization of what each person needs, and richer meeting summaries.

Built With

• adk
• docker
• java
• nuxt.js
• postgresql
• springboot