Aaron Thomas

About Aaron Thomas

Product-focused AI /ML/ Data Science engineering lead working at the intersection of machine learning, robotics, and large-scale autonomous systems. I’ve spent the last several years building products and leading teams working on autonomy evaluation, simulation, and real-world performance measurement for self-driving systems at scale. My focus is on turning complex real-world behavior into measurable systems- Building large-scale evaluation frameworks for autonomy safety, reliability, and operational performance, translating real-world fleet behavior into deployment and launch-readiness signals - Developing state-of-the-art simulation and scenario mining systems for long-tail and safety-critical edge cases using CLIP-based models, vision-language models (VLMs), and multimodal foundation models for scenario discovery and test generation - Designing decision-grade metrics that connect model behavior, perception outputs, and system-level performance to real-world autonomy outcomes - Working across engineering, product, and operations to define launch readiness criteria, guide autonomy expansion decisions, and translate evaluation signals into deployment actions Previously focused on commercial marketplace optimization, routing/dispatch, pricing and predictive modeling. More recently, my work has shifted heavily into autonomy evaluation and physical AI systems—bridging ML, robotics, and real-world deployment decisions. Expertise in: Machine Learning, Robotics systems at scale, Autonomy Evaluation, Simulation, Routing/Dispatch, Optimization, Safety & Reliability Metrics, Statistical Inference, Insight Analytics, Executive-level communication I enjoy working on hard systems problems where decisions have real-world impact and where measurement, not just modeling, drives outcomes.

Experience

1. Data Science Ai Ml Technical Lead

   Cruise

   Oct 2022 — Present

   I design and build autonomy evaluation and simulation systems used to assess safety, reliability, and operational performance of large-scale autonomous vehicle systems.In parallel, I provide technical leadership to a data science team working on evaluation frameworks, scenario mining, and predictive modeling for autonomy systems.Key areas of Work- Designing predictive evaluation frameworks for autonomy and marketplace KPIs across operating domains- Building large-scale scenario mining systems (including CLIP-based and multimodal approaches) to surface rare and safety-critical driving situations and build accurate and wide-spanning tests- Defining safety and performance metrics used in launch readiness and expansion decisions- Partnering with engineering and product teams to translate autonomy behavior into decision-grade signals- Building system to enable scalable human-in-the-loop labeling, precision/recall benchmarking, golden dataset generation, and evaluate metric drift of key data miners.This role sits at the intersection of ML systems, robotics evaluation, and real-world deployment decisions.

Education

• Princeton High School

  High School Diploma, Various Fields

  2007 — 2011

• Massachusetts Institute of Technology

  BS, Aerospace Engineering and Computer Science

  2011 — 2015

Skills

• Solidworks
• Javascript
• Html 5
• Spanish
• Cad
• Matlab
• Java
• Python
• Css3
• D3.js