Psytrade
I designed and built a research and Trading 212 Practice workspace, combining rules-based experiments with AI-assisted research and explicit operational controls.
Product and engineering ownership
I independently planned, architected and built the project, including its interface, research workflows and broker integrations.
Controlled experimentation
Matched rules-only and AI-assisted Practice experiments allow research comparisons. AI advice is bound to a defined protocol, while cash, ownership and execution controls remain explicit.
Reliability and auditability
The design includes reconciliation, bounded requests, cost controls, persistent audit records and recovery paths that avoid blindly repeating an uncertain order submission.
Current boundary
This is a research and Practice project. Real-money trading is not implemented, and the portfolio makes no claim of investment performance. This exhibit uses static descriptions and illustrative graphics, with no live broker connection.