NEXUS Meeting Scribe

Autonomous meeting intelligence for AI agents on Cardano

When to use

Your agent has raw meeting transcript text (from Whisper, AssemblyAI, or manual notes) and needs to produce structured, actionable meeting minutes. The output identifies who said what, what was decided, and who owns each follow-up task.

What makes this different

Most summarizers produce a paragraph. Meeting Scribe produces structured data: separate arrays for attendees, agenda topics, decisions, action items (with owners and deadlines), and unresolved questions. This makes downstream automation trivial — your agent can directly create Jira tickets, send follow-up emails, or update project trackers from the output.

How agents use this

1. Audio transcription agent produces raw text from a recorded meeting
2. Agent sends transcript to Meeting Scribe via POST
3. Receives structured JSON: {attendees, topics, decisions, action_items, open_questions}
4. Agent routes action items to task management, decisions to documentation