Vortraq (@vortraq) header
Vortraq (@vortraq) profile picture

Algorithmic trading research, backtesting, validation & execution. Train machine learning on proven-causal data — verified field by field, not assumed.

Joined May 2026
Photo by Vortraq (@vortraq)
Most trading systems optimize for prediction. Vortraq optimizes for validation. Composable strategies · structural execution pipelines · confidence over certainty · every layer auditable. Built for traders who want to understand and own their edge — not rent a black box.
8
1
20
228,806
Updated version now live. Previously encountered velotrade ping issue now resolved, turns out that velotrade requires systematic re authentication
96
Hyperliquid end point is now setup and operational
191
The easiest strategy to build is one that works on historical data. The hard part is proving it still works on data it has never seen before. Walk-forward testing, out-of-sample validation and realistic fees tend to change the conversation very quickly.
1
359
Observed an unusual scaling issue during long-run testing: Old signal → stable New signal → stable Old + New → severe degradation Initial assumption: detector cost. Actual root cause: repeated SHA256 hashing in shared snapshot plumbing. Interesting reminder: interactions between components often dominate individual feature cost.
3
7
505