Engineering + Statistical Analytics for Sovereign ML & Decision Systems
Founded by Guang Ouyang — Ex-Google Data Scientist (PhD Statistics)
Production ML Systems · Predictive Modeling · Analytical Infrastructure · Experimental Design · Incrementality Analysis
Axiom Data Lab aims to build sovereign machine learning and analytical systems designed for long-term analytical independence and operational freedom.
Infrastructure sovereignty means maintaining long-term analytical independence across the full machine learning lifecycle:
Rather than relying heavily on fragmented third-party platforms that may introduce API limits, escalating costs, infrastructure constraints, and long-term dependency, Axiom Data Lab prioritizes vertically integrated systems spanning data collection, validation, modeling, experimentation, and research automation.
Sovereignty also means increasing operational freedom: engineering and analytical systems should remain tools that serve business objectives and human decision-making rather than systems that continuously increase operational overhead, maintenance burden, and organizational dependency as complexity grows.
Axiom Data Lab views AI-assisted automation as a powerful enabler of operational freedom: reducing repetitive operational work, improving long-term maintainability, and allowing machine learning systems to become increasingly self-sustaining over time instead of demanding endless manual servicing.
Through building internal sovereign ML and predictive decision infrastructure for large-scale investment research across U.S. equities and ETF markets, Axiom Data Lab has experienced the operational leverage, research flexibility, and long-term freedom created by vertically integrated systems designed around sovereignty, maintainability, and AI-assisted automation. This motivates us to collaborate with external AI and data-driven teams to help apply these same principles to their own production ML and analytical infrastructure.
Internally, Axiom Data Lab is building vertically integrated data and machine learning infrastructure designed for large-scale predictive modeling research, analytical discovery, and decision support systems.
A core priority is infrastructure sovereignty across the full data lifecycle — including data collection, ingestion, normalization, validation, feature engineering, modeling, and research workflows — enabling greater long-term research flexibility, reliability, and infrastructure independence.
Current flagship research focuses on automated detection of long-horizon high-return pre-breakout setups across the entire U.S. equities and ETF market (10,000+ securities)
The broader objective is to develop infrastructure-sovereign analytical systems capable of supporting predictive research and decision systems across multiple industries beyond financial markets.
Drawing from the experience of building internal sovereign ML and analytical infrastructure, Axiom Data Lab collaborates with AI and data-driven teams to improve production ML reliability, predictive analytics quality, experimentation systems, infrastructure architecture, and long-term operational efficiency.
Services include ML system audits, metrics and evaluation design, analytical validation, infrastructure redesign, anomaly detection systems, workflow automation, and implementation support.
Engagements focus on helping organizations reduce operational fragility, improve infrastructure maintainability, and build systems that create long-term analytical leverage rather than growing maintenance overhead.
Guang Ouyang is a former Google Data Scientist with 9+ years of experience spanning:
Previous work contributed to production ML systems associated with approximately $400M annual YouTube Ads revenue impact.
Axiom Data Lab is interested in conversations with engineers, researchers, founders, and organizations exploring machine learning infrastructure, analytical systems, predictive modeling, automation, and long-term operational reliability.
Some discussions may lead to consulting, implementation support, infrastructure audits, or longer-term collaboration. Others may simply involve exchanging ideas, discussing technical challenges, or exploring research directions related to sovereign ML systems and analytical infrastructure.
The goal is to encourage thoughtful technical conversations first — especially around difficult real-world problems involving scale, reliability, infrastructure durability, experimentation, and truth discovery under uncertainty.
📩 guang@axiomdatalab.com