SynthIQ

SynthIQ turns scattered research into clear insights. It ingests sources, summarizes key ideas, and visualizes them as TL;DRs and knowledge graphs — all powered by Cloud Run.

Comment

Inspiration

We built SynthIQ to solve one problem: information overload. Researchers, founders, and analysts spend hours connecting scattered data. SynthIQ automates that process—summarizing insights in seconds.

What It Does

SynthIQ ingests web pages, PDFs, or text and produces:

  • A concise TL;DR summary
  • Thematic clusters of insights
  • FAQs for quick comprehension
  • Interactive knowledge graphs visualized in Mermaid format

All of this is displayed in a sleek, dark-mode dashboard.

How We Built It

  • Frontend: React + TypeScript + Vite + TailwindCSS + Framer Motion
  • Backend: Modular FastAPI microservices — Orchestrator, Ingestor, Summarizer, Viz
  • Infra: Docker Compose for local dev, Google Cloud Run for deployment
  • Integration: Async APIs using HTTPX and Uvicorn workers

Challenges

  • Managing async communication between multiple FastAPI services
  • Handling rate limits without Redis or Celery
  • Keeping the UI responsive during multi-stage pipeline execution

Accomplishments

  • A fully functional, production-ready monorepo built in 48 hours
  • Modular architecture ready for AI model integration
  • Real-time progress tracking and health monitoring
  • Cloud-native deployment compatible with Cloud Run

What We Learned

  • How to design scalable multi-service systems with minimal overhead
  • The importance of clarity in both UI and orchestration logic
  • That clean architecture accelerates iteration under hackathon pressure

What’s Next

  • Add Gemini 1.5 Pro for advanced summarization
  • Introduce user logins and team workspaces
  • Support ingesting Notion, Drive, and YouTube content
  • Launch SynthIQ Cloud — a collaborative research hub powered by Google Cloud

Built With

Share this project:

Updates