MindStream Mind Map Example
GitHub Repo Devpost (Gemini Hackathon)

MindStream: Accessible Mind Maps from Meetings — Powered by Agentic AI

Built at GemiHacks 2025 in 12 hours | Gemini 1.5 + Model Component Protocol (MCP)

Elevator Pitch

MindStream turns messy meeting transcripts into clean, visual mind maps using agentic AI. Designed for neurodivergent users, people with ADHD, and those who are hard of hearing, it helps everyone quickly absorb what matters.

Why MindStream?

  • Meetings can be overwhelming for people with ADHD, cognitive disabilities, or hearing impairments.
  • Traditional transcripts are long and inaccessible; static summaries lose structure.
  • MindStream uses generative AI to transform transcripts into visual, hierarchical mind maps that are easy to follow and act on.

How It Works

  • Classic Mode: One-shot prompt to Gemini 1.5 Flash, output parsed as a JSON mind map.
  • Agentic Mode: Gemini acts as an agent, calling tools (extract_structure, merge_maps, agent_memory) to build deeper, more structured maps for long or complex transcripts.
  • Speech-to-Text: Collects speech input and converts it to text for mind map generation (by Vignesh Sundararajan).
MindStream Architecture Diagram
Architecture of the MindStream application.

Technical Stack

  • AI Layer: Gemini 1.5 Flash, MCP for tool-calling, JSON schema endpoints
  • Frontend: React, TypeScript, TailwindCSS, React Flow
  • Backend: FastAPI (Python), MongoDB Atlas, MCP tool schema serving, LangChain fallback

Key Features

  • Dual mode: Classic prompt or advanced agentic reasoning
  • Tool-callable Gemini agent (extract, merge, memory)
  • Neuroinclusive UI: Visual mind maps instead of flat transcripts
  • Optional memory: Persist and evolve maps across sessions
  • Robust fallback parsing for malformed outputs

Demo Use Case

Upload a 30-minute meeting transcript. MindStream breaks it down into branches like "Launch Plan", "Frontend Tasks", "Backend Issues", and "Post-Launch Priorities", mapping relationships visually for fast review and action items.

What's Next

  • Voice → transcript → mind map pipeline
  • Persistent vector memory for smarter long-term agents
  • Editable mind maps with LLM-guided refinement
  • Export mind maps to Notion, PDF, or task systems

Team & Credit

  • Sujan Katari: Agent design, Backend, MCP integration
  • Ayaan Khan: Frontend, UI/UX Design, Graph Visualization
  • Nicholas Gonzalez: Backend, MongoDB Atlas AWS Integration, FastAPI Integration
  • Vignesh Sundararajan: Architecture mindmap, speech-to-text pipeline
  • Gemini 1.5 (Google Generative AI)
  • Powered by MCP, inspired by LangChain & LangGraph

MindStream doesn't just summarize — it understands, organizes, and visualizes complex conversations. By turning transcripts into mind maps, it makes meetings more accessible, actionable, and inclusive for everyone.