See If We Match
An agentic RAG recruiter assistant. Paste a job description and it rates every requirement Strong, Partial or Missing, backed by an exact quote and a link to the evidence. Lightweight by design: no database, no agent framework, just JSON files and one serverless function on Vercel.
Read more
Not every agent needs a heavy stack. DropOperator uses LangGraph, a Pinecone vector database and Supabase because it stores plans, calendars and user data. This agent only has to search one portfolio, so its whole knowledge base is two JSON files of about 1 MB: 71 text chunks and their embeddings. They load into memory when a request comes in, and a few lines of Python compare them by cosine similarity.
That makes it a smaller, more sustainable agent: two Python packages (FastAPI and OpenAI), no database to host or pay for, nothing running between visits, and under 1¢ per job check. When the site text changes, one script rebuilds the JSON. The trade-off is clear too: with thousands of documents, frequent updates or per-user data, I'd switch to a vector database.
I built it with Claude Code as my AI pair programmer.
