Algorithmic trading - the art of building algorithms to predict and profit from financial markets - could be one of the hardest and most lucrative data science problems in the world. Solving it is applied AI at its finest.

​

At Blackrose, we're not serving up adverts, worrying about user acquisition metrics, or building software to sell to faceless third parties. Everything we build is for our own use, and we use what we build to trade over $100+ million in assets with virtually zero human intervention.

Trading provides a unique set of challenges for machine learning research and software engineering. Datasets contain trillions of noisy observations. Data can stream faster than most software can process. Markets constantly change and adapt, causing patterns to fade and render existing models useless.

​

To keep us at the top of our game, we rely on quants that use a powerful research cluster to look for subtle patterns in petabytes of market data and devs that build meticulously engineered systems to execute trade decisions faster than the blink of an eye.

​

Rather than compete on speed to ferret out prices a nanosecond before anyone else, our edge lies in the superior predictive power of our mathematical models built using bespoke machine learning techniques and tuned on massive datasets.

​

Our bet is that a sophisticated, intelligent, and adaptive trading system will outperform a simpler or faster one.