R&D of AI & machine learning (“ML”) software for financial markets
- Investing via better performance
- Lower fees, and wider access
Perception of technology-driven investing by increasing transparency
Gather Data: Quantitative & qualitative
Collect, organize, tag, store data: textual (e.g., transcripts, filings, news articles, analyst opinions), company fundamentals & technicals, market structure, macroeconomic data, third party forecasts, alternative data.
Train Models: AI/Machine learning models
Using methods ranging from the latest natural language processing methods in deep learning to regression to tree-based learners, we construct a variety of models that are optimized to predict future performance.
Build Trader: Reinforcement Learning (RL*)
Using forecasts from our asset models, along with a variety of other market and historical data, use the latest reinforcement learning methodologies to train an agent on how to optimally buy, sell, or hold a security.
- 24+ months of development on deep network technology
- Generate 3,000+ US stock predictions daily
- Beta-testing RL agent to automate trading
- Use proficiency to build bespoke investment products & capabilities