Show HN: Aurca AI – Find Mispriced Event Contracts on Prediction Markets (aurca.ai)

Aurca AI is a prediction‑market analytics tool that runs probabilistic models against live event contracts to flag mispricings traders might exploit. The interface shows model-implied probabilities (e.g., “Highest Temperature in NYC on 25OCT31: 62–63° → 44%”) alongside current market prices quoted in cents (market price 1¢ = $0.01). Listings are updated in real time (minutes–hours), covering weather, gas prices and TSA check‑in volumes. Several clear gaps appear: NYC’s model says ~44% but the market price is 1¢ (massive divergence), Miami shows 62% vs 26¢, MDW 53% vs 31¢, while Philadelphia is nearly certain (100%) and priced at 95¢—examples that illustrate where model vs market beliefs differ materially. For the AI/ML community this is a real‑world deployment of probabilistic forecasting and decision support: models produce calibrated probability estimates that can be compared to market-implied probabilities to identify arbitrage, inform market‑making, or drive automated trading. Key technical implications include the need for rigorous calibration (Brier/log loss), robust real‑time data pipelines, latency constraints, and accounting for transaction costs, liquidity and market impact. The project clarifies how forecast accuracy, update frequency and model risk translate directly into tradable opportunities—and highlights practical challenges like overfitting to historical data and operational risk in production trading systems.