Karma Farming Autobot
A Reddit trend intelligence platform that finds what is gaining attention and explains why, with traceable metrics.
On this page
- Ended
- Ongoing
- Type
- project
Built with
Overview
A personal Reddit trend intelligence and content-assistance platform for Indian and global communities. It finds what is gaining attention — text, memes, images, video, news — explains why with traceable metrics, surfaces opportunities for genuine contributions, and helps draft original, rule-compliant content.
Despite the name, it is not a karma farm: no vote manipulation, no automated or bulk posting, no scraping. Authorized Reddit API only, and human approval for everything.
How it works
A personal Reddit trend-intelligence and content-assistance tool: it finds what's gaining attention — text, memes, images, video, news — explains why with traceable metrics, surfaces opportunities for genuine contributions, and drafts original, rule-compliant content that the user reviews and posts themselves. No automated posting, no vote manipulation, no scraping — authorized Reddit API only, human approval for everything, despite the name.
Stack: Next.js/TypeScript/Tailwind/shadcn/Recharts on the frontend; FastAPI/Pydantic/SQLAlchemy/Alembic/LangGraph on the backend; PostgreSQL with pgvector for retrieval, Redis + Celery for background jobs; Docker Compose and GitHub Actions for the rest.
Challenges
Phase 0 (the ingestion layer) is blocked, and said so rather than worked around. It's written against the official, authorized Reddit API, and is waiting on Reddit's own API approval — not swapped for scraping to unblock progress. Phase 1, the foundation underneath it, is built regardless, which is the honest way to make progress on a project with an external dependency outside your control.
Results
Phase 1 — the foundation — is built. Phase 0 is waiting on Reddit API approval, which is the current blocker: the ingestion layer is written against the official API and will not run on anything else.