I built a lightweight market intelligence layer for Kalshi prediction markets.

Instead of scraping pages, it intercepts Kalshi’s APIs to track: • Stateful volume trends (rising / falling) • Liquidity depth • Basic sentiment signals

It also includes a BYOK AI adapter so users can generate short trader briefs using their own LLM key (ChatGPT, Claude, Gemini, etc.). Designed to run on very low resources (fits Apify’s 1GB free tier).

I built this as part of the Apify $1M Challenge and would love feedback from people interested in prediction markets, market microstructure, or quant-style tooling.