Social Spacefeed

Inspiration

Social Spacefeed was inspired by real-time mission dashboards and social timelines — the goal is to make high-frequency telemetry human-friendly by surfacing concise AI summaries and visuals so teams and the public can quickly understand what matters.

What it does

Ingests live telemetry streams.

• Persists telemetry to MongoDB and serves it via a backend API.
• Runs an AI summarizer that generates short natural-language summaries and visual artifacts from telemetry.
• Pushes real-time updates to connected clients over WebSockets so the frontend shows a live feed.

How we built it

Backend: Node.js + TypeScript, Express-style API and WebSocket real-time layer.

• Frontend: React + TypeScript, components for live feeds, channels and captions.
• Data store: MongoDB for telemetry and metadata.
• AI: separate service that processes stored telemetry and writes summaries/visuals back to the backend.
• Infrastructure: Docker + docker-compose files (scaffolded) to run multiple services locally.

Challenges we ran into

• Handling high-throughput telemetry while keeping the UI responsive.
• Designing schema and indices for time-series-like telemetry in MongoDB.
• Orchestrating multiple services (ingestion, API, AI summarizer) for local development.

Accomplishments that we're proud of

A working end-to-end prototype that shows telemetry → AI summary → live UI updates Clear separation of concerns so services can scale independently

What we learned

• Practical trade-offs when turning high-volume telemetry into concise user-facing information.
• How to structure a monorepo with npm workspaces for frontend/backend workflows. ## What's next for Social Spacefeed
• Improve AI summarization quality and evaluation metrics.
• Add authentication and per-channel permissions.
• Add end-to-end tests and CI deployment pipelines.

Built With

• javascript
• phyton
• typescript