Experience

A timeline of my professional journey and key achievements.

Interview Kickstart Current

Head of AI/ML & Agentic Delivery

Jun 2024 — Present
  • Designed, coded, and shipped a production agentic AI evaluation system from scratch (FastAPI + AssemblyAI + Claude + structured-rubric LLM scoring). Processes 100+ live coaching sessions weekly; drove 22% instructor performance lift and 4.71/5.0 NPS across 400+ MAANG-level instructors.
  • Lead cross-functional delivery team including FAANG subject-matter expert working groups in agentic AI; coordinate architecture reviews, evaluation standards, and curriculum roadmap.
  • Built a DeepEval-based LLM evaluation suite with instructor-sourced golden test cases, LLM-as-judge scoring, and production quality gates — directly addressing the "ships but can't measure" failure mode.
  • Deployed Open Brain (OB1), a production Model Context Protocol (MCP) knowledge-base server on Supabase: integrates Claude with Drive, Gmail, Calendar, and a seven-category knowledge schema. Working MCP implementation predating broad enterprise adoption.
  • Researched, specified, designed, built, and tested PACCA v2.2.0, a public multi-agent healthcare prior-authorization platform: PRD + SDD applying four agentic design traditions across FastAPI/ChromaDB/PostgreSQL/React, plus HIPAA compliance docs.
  • 61.4% FAANG placement rate across 10,000+ graduates vs. 5% industry baseline (2024 internal data); Fortune 500 and federal government clients.

Omdena

Generative AI Product Manager → part-time

Jan 2024 — Jan 2025
  • Led a 71-person distributed GenAI program across 16 time zones with no formal authority — coordinated POC architecture, customer vision, and full product lifecycle for a film-industry GenAI app adopted by industry stakeholders.
  • Operated through influence and technical credibility rather than reporting structure — pattern applicable to matrixed enterprise AI programs at scale.

Stealth GenAI Startup

Principal Performance Architect → advisory, under NDA

Oct 2023 — Jan 2025
  • Designed production ML architecture for high-concurrency vector similarity search and LangGraph-based multi-agent orchestration — patterns now standard across the agentic AI ecosystem.
  • Implemented LLM evaluation pipelines with version control, A/B testing, quality gates, and observability. Designed automated prompt-engineering workflows to scale agent behavior tuning across use cases.
  • Led technical solutioning for LLM productionization: architecture selection, dependency mapping, risk mitigation, and sequencing under latency and cost constraints.

Microsoft

Principal Technical Program Manager — AI

Jul 2021 — Mar 2023
  • Owned $7M program budget across an AI initiative portfolio; drove planning, forecasting, and delivery accountability across 10+ engineering and research teams in a highly matrixed organization.
  • Drove high-priority AI programs across engineering, research, product, compliance, and legal: planning, forecasting, scheduling, dependency management.
  • Identified and championed early-stage AI capabilities (LLMs, pre-mainstream GenAI features) ahead of enterprise adoption; translated emerging research into product roadmaps and infrastructure to ship at enterprise scale.
  • Led model performance evaluation for AI features: offline benchmarks, online A/B experiment frameworks, and quality KPIs for LLM-powered enterprise products.
  • Communicated program goals, status, and trade-offs to senior executives and technical audiences; defined roadmap and long-term strategy in partnership with engineering leadership.

Medalogix

Principal Data Scientist / Senior Director

Nov 2020 — May 2021
  • Owned production AI/ML system delivery end-to-end in a regulated healthcare context: NLP modeling, inference pipelines, data governance, and model lifecycle management under direct business accountability.
  • Directed a team of 6 data scientists and ML engineers to define, build, and validate production predictive models for critical healthcare outcomes in a regulated clinical environment.
  • Performed technical AI due diligence contributing to a $250M acquisition — evaluated scalability, model quality, and long-term system maintainability.

Trilogy Education (acquired by 2U)

Lead AI Curriculum Architect / Product Manager

Jun 2019 — Nov 2020
  • Delivered a $70M cross-functional technical program across six R1 universities (Columbia, Penn, UC Berkeley, UT Austin, UNC, Toronto); managed 100+ stakeholders across engineering, content, operations, legal, and executive leadership.
  • Led a cross-functional team of 8 (engineers, curriculum architects, operations) across program delivery; managed alignment across 100+ contacts at six R1 universities.

Deep Advisors (Independent Practice) Current

Deep Learning Consultant → part-time advisory

Jan 2018 — Present
  • Delivered 20+ end-to-end AI/ML programs for Fortune 500, federal, healthcare, and R1 university clients — NLP, computer vision, recommendation systems, and GenAI/agentic systems.
  • Translated business goals into executable ML program plans: scope definition, dependency mapping, risk mitigation, KPI frameworks for use cases with significant ambiguity and regulatory constraints.

Oracle Corporation

Senior Technical Director — Data Warehousing / AI

Jan 1997 — Dec 1999
  • Led a 16-person engineering team to rebuild Amazon's middle-tier e-commerce platform on Oracle infrastructure — managed a $1.25M program budget, milestones, and Oracle/Amazon dependencies.
  • Architectural patterns from that system (real-time behavioral signals, intent modeling, large-scale distributed inference, continuous model improvement) are direct predecessors of modern ads ranking and agentic retrieval systems.
  • During this engagement: sole inventor of US Patent 6,850,988 — clickstream personalization algorithm foundational to modern recommendation and behavioral-signal systems.

Earlier Career

Master Technologist · VP Technical Services · Strategist · Graduate Faculty

Jan 1999 — Dec 2018
  • Master Technologist, Hewlett-Packard — IC-track equivalent to Principal/Distinguished Engineer; global consumer data warehouse architecture.
  • VP Technical Services, VisionCompass — 125-person offshore software organization.
  • Strategist / Product Manager, BPM Northwest — AI/ML digital transformation consulting, C-suite advisory.
  • Graduate faculty: Walden, Capella, Kaplan (2009 – 2018).

Skills Matrix

An honest assessment of where I excel, where I'm growing, and where I have gaps.

✓

Strong

16
Multi-agent LLM systems Agentic AI delivery & governance LLM evaluation (LLM-as-judge, DeepEval) Retrieval-Augmented Generation (RAG) Model Context Protocol (MCP) LangGraph LangChain Claude API (Anthropic) Python (FastAPI, async) Production ML engineering Cross-functional delivery (matrixed orgs) Healthcare-regulated AI (HIPAA) Senior IC / staff leadership 0→1 product delivery Program & budget ownership ($7M+) Stakeholder & executive communication
○

Moderate

16
PyTorch TensorFlow scikit-learn Hugging Face GPT-4o / OpenAI APIs Gemini Llama / open-weights models TypeScript / React PostgreSQL ChromaDB Docker AWS (SageMaker, Bedrock, EMR, Kinesis) Azure AI Foundry Google Cloud (Vertex AI) Databricks / PySpark Snowflake
⚠

Gaps

5
Distributed training at scale (pre-train / large fine-tune) Low-level CUDA / kernel optimization iOS / mobile native development Graph databases (Neo4j) at production scale On-device ML / edge inference deployment