Trade4Me - Trading Project

Trade for me.

Global macro trends provide us with invaluable insights into the potential future. My objective is to diligently monitor these trends and strive to interpret them with utmost accuracy. The purpose of this project is to develop practical automations that facilitate efficient market observation and enable prompt decision-making. In the final stages, I aim to leverage machine learning to automatically execute trade decisions.

Three-level Approach

Global Macro level

Get macro economics factors, companies quarter reports to predict trends on high-level and for long-term.

Company level

Open/Close statistics on market for tech-analysis and med-term horizon prediction.

Daily-trading level

5min trading data and news for inraday trading.

Daily-trading level

News Parser

Daily News for intraday trading - collected info for fast decision-making. From various source get text info, store it, then using ML get summary and sentiment analysis. Based on user preferences sort all news and send as telegram notification only top of the list.

News Parser scheme

TextSummary

Dedicated service for Summarization and Sentiment of the text. For it, I use dockerized FastAPI server with simple Gradio interface. Additionally, there are lazy loading for ML models to release resources and verification to prevent bots from constantly using this service. API available through basic authentication.

Text Summary interface