THE FIRM
We build machines that trade.
ATRX is a quantitative technology firm. We design, build, and operate AI native trading infrastructure that reasons about markets at machine speed, adapts to regime shifts in real time, and compounds intelligence from every decision it makes.
We are not a hedge fund. We are not a trading bot company. We are a technology firm that happens to operate in financial markets because markets are the most demanding environment on earth for autonomous decision systems.
If your system works here, it works anywhere.
TWO ARMS
Proprietary capital meets scalable infrastructure.
ARM I: PROPRIETARY TRADING
ATRX trades capital directly across global financial markets. Forex. Metals. Crypto. Indices. Commodities. The system operates autonomously, 24 hours a day, five days a week, without human intervention. Every trade is executed by an AI agent network operating under strict, hard coded risk architecture that no model can override.
This is not paper trading. This is not simulation. This is real capital, real markets, real consequences. The system has been live since November 2025.
ARM II: TRADING AS A SERVICE
The same intelligence that powers our proprietary trading is packaged into a downloadable agent that runs on the customer machine. The customer installs it. Connects their broker. Sets their risk limits. The agent starts trading.
Intelligence flows from ATRX cloud servers. Execution happens locally. Broker credentials never leave the customer device. This is infrastructure, not a subscription to someone else's trades. Every customer runs the same caliber of technology that institutional desks spend millions building internally.
THE STACK
Five pillars. Built from first principles.
CAUSAL MACRO INTELLIGENCE
Most trading systems treat macro data as noise or as simple correlation signals. ATRX maps causal pathways showing how global events transmit to specific assets. Rather than relying on rigid historical correlations, the system embeds economic reasoning to evaluate whether relationships hold in novel situations and shifting regimes.
MULTI MODEL AI REASONING
Multiple frontier AI systems are deployed simultaneously, each assigned to strategy types where they perform best. Every decision is attributed and scored. This creates a continuous, real world benchmark of which AI architectures excel at which specific financial reasoning tasks.
ADAPTIVE STRATEGY LIBRARY
We maintain a dynamic library of researched alpha strategies, each built on distinct market phenomena. The deployment engine activates and suspends strategies based on prevailing market conditions. This allows new approaches to integrate seamlessly without disrupting the existing ecosystem.
INSTITUTIONAL RISK ARCHITECTURE
Multiple layers of quantitative control gate every trade. These are absolute, hard coded constraints that no AI model can override, covering position sizing, exposure limits, asset blocking, and crisis detection.
COMPOUNDING INTELLIGENCE
Every decision across the system is recorded with full attribution. This data feeds our continuous learning loops, allowing the infrastructure to adapt, optimize parameters, and detect strategy degradation before it becomes drawdown. The longer ATRX runs, the more capable it becomes.
ARCHITECTURE
Intelligence flows down. Data flows up.
LAYER 1: THE INTELLIGENCE CORE
The engine. Computes the market state, evaluates macro conditions, and broadcasts validated strategy signals. Runs on ATRX cloud infrastructure, computed once, and served to every agent instance.
LAYER 2: THE AGENT NETWORK
The deployment. An isolated swarm of autonomous agents, each with a single responsibility, from diagnosis to execution. The network connects to the customer broker through a proprietary gateway, operating locally under strict risk guardrails.
LAYER 3: THE INTERFACE LAYER
The experience. A premium command center showing real time profit and loss, live regime visualization, strategy health tracking, and decision attribution. Every action the system takes is visible and traceable.
ORIGIN
Built by one engineer. In the market.
ATRX was built by Timilehin Olapade, a quantitative engineer and the sole founder of the firm. Every component of the system, from the stochastic models to the multi model LLM pipelines to the risk architecture, was designed, built, and tested by a single person.
This is not a story about privilege or access. There was no lab. No incubation. No seed funding. No team of PhDs. The system was built in the market, trade by trade, solving real problems in real time, funded by the founder's own capital and validated by live performance.
Timilehin's background spans quantitative finance, software engineering, and machine learning. He is currently enrolled in the Master of Science in Financial Engineering program at WorldQuant University.
The founding story matters because it defines the company DNA. ATRX was not designed in theory and then implemented. It was built under constraint, validated under pressure, and refined through live market feedback. That discipline is embedded in every layer of the architecture.
WHERE WE ARE GOING
The first AI native quantitative technology firm built for global markets.
The firms that dominate quantitative finance today were built over decades with hundreds of millions in infrastructure investment. They are brilliant. They are also closed, concentrated, and aging.
Three shifts are creating a window:
Frontier AI models are now accessible via API. For the first time, a small team can deploy reasoning capabilities that rival institutional internal systems.
Macro data is becoming real time and democratized. Prediction markets, economic feeds, and sentiment data are now available programmatically.
Quantitative strategy alpha is decaying faster. As more participants deploy similar algorithms, edges erode quickly. The winners will be platforms that compound intelligence, not static strategies.
ATRX is positioned at the intersection of all three shifts. The technology is built. The system is live. The architecture is designed to scale across asset classes, geographies, and customer segments.
The quality of the technology dictates the outcome, and the infrastructure is ready for global scale.