Solution diagram
The challenge
Build an automated trading system with data processing, explainable decisions and remote controls.
How the solution works
Twelve Python microservices using Clean Architecture, ClickHouse and PostgreSQL. Trading connects through the T-Invest API; each decision is explained, with controls available through Telegram.
Implemented capabilities
- 12 microservices with separate responsibilities
- T-Invest API, ClickHouse and PostgreSQL
- Decision explanations and Telegram controls
How it helps the business
- Separated services support independent changes to data processing and application logic.
- Explanations provide evidence for reviewing algorithm behaviour.
- Telegram offers another way to control the system. Automation does not guarantee returns.
Where this fits your project
For products with data streams, algorithmic decisions and a need to explain those decisions.