Explainable, Adaptive AI
for Agentic Finance
AI agents that learn in real time, explain every decision, and actively reduce uncertainty. Built on peer-reviewed Active Inference research.
Building in public. Validated by research.
260+ Live Agents
Learning agents processing real market data in production. Active Inference-powered agents that adapt to market regimes without retraining.
Peer-Reviewed Science
Published in Entropy (MDPI, 2026). Introducing the Entropic Sharpe Ratio: return per bit of information processed.
Read the paper →Aspen Institute Partnership
Strategic agreement with the Aspen Institute to co-host the Agentic Finance Summit (June 2026). Defining the standards for AI-driven investment management.
Active Inference nests both Black-Scholes-Merton and Discounted Cash Flow as special cases — adding real-time Bayesian learning and intrinsic uncertainty reduction.
A new kind of investment AI
AI agents powered by Active Inference — to our knowledge, the only framework that delivers real-time learning and full explainability.
Explainable
Every decision has a full audit trail. Trace actions back to beliefs, observations, and preferences. Designed for the EU AI Act era — full explainability for high-risk AI systems.
Adaptive
Real-time Bayesian learning without retraining. Agents update continuously as new data arrives.
Discovery-Driven
Agents actively seek information, balancing known strategies with exploration of new signals.
From research to production
Market Modeling
Assembling best-performing agents into portfolios. IP investigation.
Scientific Discovery
Automated strategy discovery via SciAgents methodology. IP filings.
Adam Interface
1M agents modeling 10K securities. Natural language portfolio construction.
Instrument Design
New instruments from agentic predictions and actions.
ETF Launch
Design and launch agentic ETFs. Subject to regulatory approvals.
Market Modeling
Assembling best-performing agents into portfolios. IP investigation.
Scientific Discovery
Automated strategy discovery via SciAgents methodology. IP filings.
Adam Interface
1M agents modeling 10K securities. Natural language portfolio construction.
Instrument Design
New instruments from agentic predictions and actions.
ETF Launch
Design and launch agentic ETFs. Subject to regulatory approvals.
12-Month Targets
7 FTEs · Investor Product Release · 5 Family Office Clients · IP Filings
Built by operators, researchers, and pioneers
Repeat entrepreneurs, AI researchers, and investment professionals.
Repeat founders with successful exits. Techstars alumni. Experience directing international accelerator programs and managing portfolio operations at scale.
Pioneers in multi-agent AI, computational finance, and stochastic inference. Published researchers with decades of academic and applied work.
Investment professionals with experience across private equity, venture capital, and institutional asset management totaling $70B+ AUM advisory.
Board & Advisors
Thomas Farb-Horch
Advisor
7 multi-billion dollar exits including creating FICO scores. GP, private equity & VC.
Vadim Toader
Advisor
AI-exited founder ($150M+). Forbes 30 Under 30. Ex-Bain, Google ML. M.Eng Oxford.
Michael I. Miller, Ph.D
Scientific Advisor
Chairman, Johns Hopkins Bioengineering. Internationally recognized expert in stochastic inference.
Fernando Reyes
Finance Advisor
Rebuilt hedge fund factors (Yale/AQR). Predictive asset pricing models. Research Associate, Yale Center for International Finance.
Stay Ahead of Agentic Finance
Join our research updates list for quarterly insights on Active Inference, agentic investing, and our progress.