Developer builds self-hosted Kalshi signal monitor after struggling to read prediction markets

A developer built an open-source tool called Trade Hunter after finding Kalshi's prediction market data difficult to interpret as a newcomer. The tool subscribes to live Kalshi WebSocket feeds, automatically tracks all contracts within multi-contract series, and surfaces unusual price or volume movements on a local dashboard in real time. A spike detector with configurable thresholds sorts signals into tiers, prioritising moves where score, price, and volume conditions all trigger simultaneously. Two AI layers sit on top of the detector: one narrates each spike with a plain-English read and confidence level, while a second analyses recent false positives and suggests threshold adjustments the user can apply in one click. The developer notes the underlying AI model does not retrain — only the detector's configuration improves over time — and has shared the writeup publicly to invite community feedback on the design choices.