AI Transformation • Applied AI • Technical Leadership

I build AI systems — and the organizations that can successfully use them.

I design, deploy, and scale applied AI systems while helping engineering organizations turn emerging AI capabilities into measurable business and developer impact.

Impact at scale

Recent enterprise AI transformation work at ZoomInfo, measured across production deployment, enablement, and adoption.

15+
Production AI agents

LLM-powered agents deployed across 10+ business organizations.

500+
Engineers enabled

Enabled globally in LLM development, prompt engineering, and Responsible AI.

95%
Enterprise AI adoption

Grew AI adoption from ~20% to 95% through deployment and enablement.

68
Enterprise AI initiatives

Led a portfolio spanning Engineering, Product, GTM, HR, Finance, Legal, and Ops.

What I do

Four disciplines that make AI work in production

Deploying AI successfully takes more than an API call. It requires architecture, evaluation, workflow design, enablement, and measurement working together.

AI Systems

Design and implementation of RAG systems, AI agents, LLM applications, evaluation frameworks, provider abstractions, and production AI architecture.

  • Retrieval-augmented generation with grounding and citations
  • Agentic workflows and tool-using LLM applications
  • Evaluation harnesses for retrieval and generation quality
  • Provider abstraction so models are replaceable, not load-bearing

AI Transformation

Turning AI experimentation into repeatable organizational capability through strategy, governance, workflows, and deployment.

  • Use-case discovery through measured production deployment
  • Reusable workflows, prompt libraries, and deployment playbooks
  • Responsible AI governance and risk frameworks
  • Executive alignment on where AI creates real value

Developer Productivity

Applying AI to engineering workflows, onboarding, enablement, knowledge access, and developer effectiveness.

  • AI-assisted engineering and knowledge retrieval
  • Onboarding and enablement systems for engineering teams
  • Embedded assistants inside real developer workflows
  • Measuring effectiveness, not just activity

Adoption & Enablement

Creating the systems, programs, measurement, and organizational behaviors necessary for teams to actually adopt AI.

  • Enablement programs for hundreds of engineers globally
  • Role-based guidance, assessments, and certification tracks
  • Adoption measurement and continuous improvement loops
  • Change management from experimentation to standard practice
Featured work

Engineering Onboarding Copilot

Evaluated RAG for grounded engineering knowledge.

Personal / portfolio project

The problem

Engineering knowledge is fragmented across documentation, runbooks, architecture guidance, operational standards, and tribal knowledge. LLMs can produce fluent answers, but enterprise knowledge systems need more than fluency.

Enterprise knowledge systems need more than fluency:

  • Relevant retrieval
  • Evidence provenance
  • Measurable quality
  • Grounded generation
  • Citations
  • Safe abstention

Technology

PythonFastAPIPydanticPostgreSQLpgvectorDockerOpenAI embeddingsRAGVector searchRerankingpytestEvaluation pipelines

Retrieval evaluation

Global corpus

Global chunk-level retrieval evaluation over the full corpus. This is the noisier, more realistic setting — and it shows the measurable effect of reranking.

Vector Vector + Reranking
Top-173.3% 93.3%
Recall@386.7% 93.3%
MRR0.867 0.967

Generation evaluation

Curated suite

Results on the current curated evaluation suite. These measure the behavior of the generation and grounding layer on a defined set of cases — they are not claims of universal model accuracy.

Grounding Accuracy
100%6/6
Citation Presence
100%4/4
Citation Validity
100%4/4
Abstention Accuracy
100%2/2
System architectureRAG pipeline
  1. Knowledge DocumentsDocs, runbooks, standards
  2. Section-Aware ChunkingPreserve structure & context
  3. EmbeddingsOpenAI embeddings
  4. PostgreSQL + pgvectorVector store
  5. Vector RetrievalSemantic nearest neighbors
  6. Candidate EvidenceBroad candidate set
  7. RerankingBest evidence, not just nearest
  8. Context AssemblyGrounding context
  9. LLM GenerationAnswer synthesis
  10. Grounding / Citation ValidationVerify against sources
  11. Answer + SourcesCited response or abstain
engineering-onboarding-copilot · demo
Question
My deployment is live and customers suddenly cannot use the product. What should the team do?
Response Grounded
Notify the incident channel, assign an incident lead, document customer impact, prioritize mitigation, stop further rollout, and determine whether rollback is safer than forward remediation.
Sources · 3
Incident EscalationRunbooks > Incident Escalation
Rollback ProceduresRunbooks > Rollback Procedures
Engineering Production HandbookHandbook > Rollback
View Source on GitHub
Contact

Building something ambitious with AI?

I'm interested in conversations around AI transformation, applied AI systems, developer productivity, AI deployment, and technical leadership.