I want a trading bot that can automatically take the “Low / High BTC” position on Polymarket every 15 minutes, analyse the outcome, and refine its own strategy as more data comes in. The core logic needs to be written in Python, but I am open to additional languages or frameworks if they will speed up development or improve long-term maintainability. Key points you should know – It will run on a desktop machine (Windows), so a simple CLI with clear logging is fine for the first version. – All decisions must be driven by machine-learning algorithms rather than fixed rules. I am especially interested in approaches that combine real-time market odds, recent price movement, and historical trade performance. – The architecture should make retraining painless: think modular data-collection scripts, a model layer I can swap or retrain, and a trading engine that simply calls the latest model. – Safety features such as position sizing limits, error handling for API downtime, and a dry-run/back-test switch are essential from day one. Deliverables 1. Well-documented Python source code (model training, prediction, and order execution). 2. README with setup steps, configuration options, and examples of how to retrain the model when new data accumulates. 3. A short demo video or live walkthrough proving the bot places at least a few successful trades on a test account. Please highlight your relevant experience building trading bots or ML-driven decision systems in your bid. I will shortlist based mainly on that practical background.