Kelvin Temperature Prediction Bot

I need a bot—internally nicknamed the “Chicken Road Prediction Bot”—that focuses exclusively on weather data. Its single job is to predict temperature, and every output must be expressed in Kelvin. No traffic flow analysis, no pedestrian tracking, and no rain or wind forecasts—temperature only.

The workflow I imagine is straightforward: the bot ingests current and historical weather information (open-source APIs are fine), runs its model, then returns near-term temperature predictions that I can call programmatically. Accuracy matters more than speed; I’m happy to discuss the modelling approach you feel suits this best, whether that’s a lightweight regression or a deeper neural network.

Deliverables should include:
• Source code with clear comments
• A simple README explaining setup and required dependencies
• An endpoint or script that returns temperature forecasts in Kelvin for a supplied location and time window
• A brief report on model accuracy and any tuning you performed

I’ll test by comparing your Kelvin outputs against live data for several cities over a 48-hour period. If you believe another step would strengthen validation, let me know—flexibility on the technical path is fine so long as the bot stays laser-focused on Kelvin temperature prediction.