Profile

I run production AI infrastructure and developer platforms at Moody's Analytics, with 17 years building systems at scale, with the last three years focused on AI: RAG pipelines, LLM evaluation frameworks, agentic workflows, and the developer platform that hundreds of engineers depend on to build, ship, and operate software. I work directly with engineering leads across business units to implement AI systems, measure their quality, and drive adoption to production. I hold a Master's in Data Science from UC Berkeley (GPA 3.81) and teach RAG, agents, and Claude to the LatinX engineering community through Techqueria workshops and my podcast.

Experience

Director, Developer & Platform Experience

2024 – Present · Moody's Analytics · San Francisco, CA (Remote)

Own and operate the developer experience platform serving multiple enterprise product teams across Moody's Analytics. Partner directly with engineering leads across business units to implement AI systems, diagnose adoption blockers, and drive platform onboarding from NPE through Canary to Production, responsible for outcomes, not handoffs. Drove adoption of AI-assisted development tools (Cursor, GitHub Copilot, Windsurf, Claude Code) and designed MCP server integrations across internal knowledge bases, API specs, and source repositories, expanding AI-native developer workflows at scale. Architected the MAP platform onboarding framework with clear NPE → Canary → Production readiness gates for Business Units adopting the Moody's Analytics Platform. Lead cross-functional execution across Platform Engineering, SRE, Security, and BU engineering teams.

AI Architect

2023 – 2024 · Moody's Analytics · New York, NY (Remote)

Designed and shipped RAG pipelines, vector search infrastructure, embedding workflows, and agentic systems for enterprise product teams, diagnosing real-world AI failure modes (retrieval mismatch, hallucination, latency) and delivering architectures teams could ship with confidence. Designed and operated the LLM evaluation framework for generative AI products at Moody's Analytics, built on Databricks, Spark, and MLflow, establishing Recall@k, Precision@k, MRR, and groundedness as standard evaluation practice adopted across AI product teams. Delivered hands-on training on embeddings, vector search, prompting, and LLM evaluation to multiple engineering organizations, transferring production AI literacy at scale across teams building enterprise AI products.

Associate Director, DevOps Architect

2015 – 2023 · Moody's Analytics · New York, NY (Remote)

Led DevOps architecture and platform engineering across a multi-cloud environment (AWS, Azure). Designed secure-by-default CI/CD pipelines, containerization strategies (Docker, Kubernetes), and infrastructure-as-code tooling (Terraform, GitHub Actions) adopted across multiple engineering teams. Drove the transition from monolithic deployment models to microservices, reducing release cycle time significantly.

Senior DevOps Engineer

2009 – 2015 · Moody's Analytics · San Francisco, CA

Built automated build pipelines, CI systems, monitoring infrastructure, and early cloud deployments supporting multiple product lines. Developed installation tooling, deployment automation, and customization frameworks, improving reliability, repeatability, and developer workflows across the organization.

Education

Master of Information and Data Science (MIDS), GPA 3.81

2021 – 2023 · UC Berkeley School of Information · Berkeley, CA

Bachelor of Science, Computer Science

California State University, Monterey Bay

Community & Speaking

Host: AI Builders: LatinX Edition, a practitioner-focused podcast teaching Claude, RAG, agents, and LLM infrastructure through conversations with practitioners shaping real-world AI adoption. 6+ episodes since September 2025, distributed on Spotify, Apple Podcasts, and Amazon Music.
Instructor at Techqueria: delivers hands-on AI workshops to LatinX engineers and early-career practitioners covering Claude, RAG systems, embeddings, agentic workflows, and LLM evaluation fundamentals. Authored open-source education materials (GenAI-Tutorial, AI Para Todos, Multi-Doc RAG with LlamaIndex) used in community workshops.
Builds with Claude Code daily, shipping this portfolio, live MCP server implementations, and edge AI workers on Cloudflare Workers as production engineering artifacts. Uses AI-assisted development as a core part of production engineering workflow, not a side experiment.
Speaker, Techqueria Tech Summit (Oct 2025): Building AI Agents for Real-World Workflows
Speaker, Oakland Tech Week (Nov 2025): AI Agents from Scratch
Speaker, Latino AI Summit (Jan 2026): RAG Chat + AI Agents
Writing: medium.com/@vhr1975

Skills

Python
TypeScript
Java
C#
C++
SQL / Spark SQL
RAG Pipelines
LLM Evaluation
Agentic Workflows
Vector Search
Embeddings
MCP / Tool Use
LangChain / LangGraph
FastAPI
React / Next.js
Node.js
AWS / Azure
Kubernetes
Docker
Terraform
GitHub Actions
Databricks / Spark
Pandas / NumPy / Scikit-learn
TensorFlow
MLflow
Cloudflare Workers
Cursor / GitHub Copilot
Claude / GPT-4o