The Future of Capital Allocation
Data-First. AI-Native. Adaptive
About Us
Malleroy Avantgarde is building the next evolution in systematic macro trading: self-improving, multi-agent intelligence engineered for institutional deployment.
We develop data-first, AI-native reinforcement learning systems that operate autonomously across global macro markets. Our mission is to transform high-dimensional data into resilient, risk-adjusted performance.
The Technology
The Agentic Market Engine
Our core platform is built on multi-agent reinforcement learning, combining deep neural networks—including recurrent architectures—with generative AI. Specialized agents coordinate through a decision mechanism that aggregates competing signals and adapts continuously to shifting market conditions.
Risk-Aware by Design
Every decision is optimized to balance return with resilience. Our models incorporate risk-sensitive objectives, including Sharpe ratio maximization and drawdown limits, ensuring performance aligns with institutional risk mandates.
Scalable Intelligence
Our modular, cloud-native architecture enables seamless horizontal scaling via microservices and containerized deployment. With support for edge inference, the system delivers low-latency execution under 100 milliseconds while remaining flexible for expansion into new asset classes and markets.