I am an AI/ML Research Scientist and Engineering Leader with 17+ years of experience innovating across statistical machine learning, Bayesian & frequentist methods, model interpretability, and modern LLM systems. I also serve as a Visiting Lecturer at UC Irvine.
Pushing the limits of how humans and AI reason through language – one commit at a time.
Oid Labs · GitHub · SlideShare
Current
- Oid Labs (Founder) Building safe, personalized assistants for Intelligent Process Automation using LLMs tailored to vertical tasks.
- LeIA: Legal Intelligent Assistance. A unified marketplace for task-specific AI assistants that automate paralegal workflows for small to mid-sized law firms.
- Lexoid: An open-source universal adapter for LLMs. Turn messy real-world data into clean, agent-ready context.
Few Highlights
H2O.ai
- LLM Robustness & Alignment Optimization: Spearheaded research and services for automated LLM evaluation with task-specific synthetic test generation via auto prompt optimization and adversarial alignment.
- SQL Sidekick: Led development of an LLM2SQL assistant enabling conversational Q&A over enterprise tabular data.
- Model-Analyzer: Conceptualized, researched, and productionized a unified interactive framework for simulation (what-if scenarios) and adversarial robustness to stress-test model behavior and limitations.
- AutoInsights: Collaborated closely with late Dr. Leland Wilkinson (Chief Scientist) to invent and launch an automated, self-service AI system that surfaces hidden data insights (auto-EDA) using natural-language narratives.
- Driverless AI MLI: Drove ML innovation and product strategy for machine learning interpretability (MLI) algorithms in H2O Driverless AI.
- Wave Image Styling Playground: Creative generative AI exploration for interactive image styling.
Before joining h2o.ai,
DataScience.com (Acquired by Oracle)
- Skater: Pioneered early enterprise model interpretability & explainability. Created and open-sourced Skater for model evaluation and opacity reduction. Watch the talk: Unboxing Model Opacity (YouTube).
Talks, Articles & Podcasts
Podcasts & Deep Dives
- Data Skeptic: Unpacking AI Hype to Reality with Kyle Polich
- Blog: State of Text-to-SQL
- ODSC: Model Evaluation in the Land of Deep Learning
Conference Talks & Workshops
- Explainable AI: Hands-on with Explainable AI at MLI Gym
- JupyterCon: Using Jupyter’s Interactiveness to Build Better Predictive Models
- Decks & Presentations: SlideShare Portfolio