I got tired of rate limits and slow responses from the official vulnerability databases while building a security tool. So, I decided to build my own solution.
I processed 25 years of CVE data (1999-2025), cleaned it for LLM context, and indexed it on a lightweight SQLite backend hosted on Render.
It returns structured JSON suitable for AI agents and security automation tools.
*Tech Stack:* Python FastAPI, Pandas (for ETL), SQLite (for search), Hugging Face (for data storage).
It is live on RapidAPI with a free tier for developers: https://rapidapi.com/cyberapiui/api/cybersec-intelligence
Would love to hear your feedback!