AI trading bot

The trader wanted to move away from manually sitting in charts to a system that itself catches entry points using a trained model, and not according to strict rules from a set of indicators

I trained a neural network on PyTorch using historical quotes and put it into a bot that collects data, calculates features, predicts and generates signals. Inference was separated from strategy in order to change the logic of inputs and outputs without retraining the model

• Trained a neural network on PyTorch using historical quotes
• Automated data collection, feature calculation and signal generation
• Separated the inference from the trading strategy to change the logic without retraining
• Implemented model retraining using fresh history
• Packaged in Docker and launched trading