Matters I'm equipped to address
Most AI disputes are not really disputes about algorithms. They are disputes about whether a system was built the way someone said it was built, whether it could ever have performed as promised, and whether the people who deployed it understood what they were deploying. Those are questions about architecture, engineering practice, and documentation — which is the work I have done for eighteen years.
Subject matter
Failed AI implementations and vendor disputes. Whether a delivered system met the specifications represented, whether the architecture could have supported the promised timeline or performance, and whether the delivery failures were foreseeable to a competent practitioner. This is the work I do in commercial advisory practice, applied to a litigation posture.
Model behavior and algorithmic decisioning. How a model reached its outputs, what the training data and design choices made likely, and whether observed disparate outcomes are attributable to model design, data selection, or deployment context.
Data governance and regulatory exposure. What data moved where, which systems and automated agents had access to it, whether controls were adequate to the represented standard, and what a reasonable practitioner would have done differently.
Technical architecture explanation. Translating machine learning systems, data pipelines, and cloud infrastructure into language a judge, jury, or arbitrator can follow without distortion.
Relevant background
Eight years as a Senior AI Architect at Google Cloud, working across financial services, healthcare, and regulated enterprise accounts where AI failures carry regulatory consequences. Prior architecture and strategy roles at Adobe and Omnicom serving Fortune 500 data organizations, and earlier at PwC.
Editor of Data Science on the Google Cloud Platform (O'Reilly Media), a practitioner's guide to building production-grade AI and ML pipelines at enterprise scale. Contributor to Google Cloud's open-source MLOps ecosystem, including AutoMLOps. M.S. in Data Science from Northwestern University, summa cum laude. Named inventor on a filed provisional patent. Founder of NoVAI, a practitioner-led AI community in Northern Virginia, and of Spartera.
The question in most AI disputes is not whether the technology is sophisticated. It is whether anyone can explain, on the record, what it actually did.
Engagement and conflicts
I take a limited number of litigation engagements and screen for conflicts before substantive discussion. Initial contact should describe the matter at a high level only — parties, jurisdiction, posture, and the technical questions in dispute — so that a conflicts check can be completed before any confidential information is exchanged.
Engagements are available on a consulting-expert or testifying-expert basis. Rates are provided on request and differ by activity: case review and analysis, report preparation, deposition, and trial testimony. Retainer required.
Related writing
The case studies below describe the commercial version of this work — the same technical analysis delivered to boards rather than to counsel.