Briefing / 01
AI research lab / Quantitative trading
Collective
intelligence.
Models that learn to coordinate.
MASSTEC is an AI research lab working on collective intelligence: populations of specialized models, trained independently with reinforcement learning and evaluated against each other in multi-agent market simulation. We apply that research to quantitative trading. Every swarm passes offline evaluation and simulation before live inference, and live swarms are coordinated across the book, each kept distinct.

Four environments / Three gates
- 01
Training
Agents trained independently on narrow objectives, composed into swarms around a mandate.
- 02
Offline
Scored on held-out market data, excluded from training. Hypothetical.
- 03
Simulation
A multi-agent market that reacts to the swarm. Hypothetical; nothing at risk.
- 04
Inference
Swarms that pass both gates, coordinated across the book.
Important information
This briefing describes MASSTEC's research and trading approach. It is not an offer or solicitation to buy or sell any security or investment service. Imagery is a conceptual illustration, not market data or results. Results from offline evaluation and simulation are hypothetical and differ from live trading.