The problem
Individual indicators and models often produce conflicting signals, making it difficult to assess quality and risk consistently.
Project 09 · Machine learning
KAI Crypto Trader
A research and decision-support platform that combines multiple models, market indicators and TradingView signals. The system assesses signals for alignment, risk and historical performance without guaranteeing profit.
Discuss a similar project →
The approach
An ML ensemble that combines signals, calculates scores, processes TradingView data and analyses performance by strategy and market condition.
The impact
A better-supported comparison of signals, more transparent risk assessment and a reproducible research environment for trading strategies.
My role
Concept, model architecture, signal logic and evaluation methodology
I connect strategy, user needs, technology and execution. The result is not an isolated demo, but a product that can be used, maintained and expanded in practice.
For this project, technical choices were continuously connected to the daily process, the people using it and the result the solution needed to deliver.
The building blocks behind KAI Crypto Trader
PythonMachine learningTradingViewBacktestingData analysis