NBCUniversal
Full-stack estimating & takeoff
A Python and React/TypeScript platform that combines document ingestion, OCR, spatial registration, and AI-assisted extraction.
Explore full-stack estimating & takeoffOmkar Chittar Sunnyvale, CA
Software Engineer / AI EngineerBuilding production AI systems.
I build production AI systems with RAG, AI agents and tool calling, and LLM evaluation—backed by MLOps and full-stack software engineering.
From knowledge discovery to workflow automation, I deliver enterprise AI from requirements through deployment and ongoing reliability.
- RAG
- Agents
- Evaluation
- MLOps
- Full-stack software
01 / Applied AI
Selected Work
Full-stack applications and AI systems for estimating, knowledge discovery, and operational support.
NBCUniversal
Legal retrieval & evaluation
Retrieval and evaluation infrastructure over 5,000+ pages of legal documents, with semantic search and automated citation validation.
Explore legal retrieval & evaluationFOX Sports
AI-powered operational monitoring
An event-driven monitoring service that surfaces ML pipeline drift alerts, exceptions, and health insights in real time.
Explore ai-powered operational monitoringFOX Sports
LLM evaluation & serving
Evaluation and regression tooling that improves retrieval quality, alongside serving optimizations that reduce per-request cost.
Explore llm evaluation & serving02 / How I build
Engineering Capabilities
RAG & Agents
Ingestion, embeddings, vector retrieval, grounded generation, tool calling, and guardrails.
LangChain · LangGraph · FAISS · Bedrock
Evaluation & Reliability
Golden sets, retrieval checks, regression harnesses, and prompt and embedding experiments.
Ragas · Retrieval evaluation · Quality regression
MLOps & Deployment
Deployment workflows, access controls, logging, monitoring, and serving cost optimization.
MLflow · Docker · CI/CD · AWS · GCP
Full-Stack & Integrations
Python backends, React/TypeScript frontends, REST APIs, and integrations with enterprise data and workflows.
Python · React · TypeScript · Flask · REST APIs
03 / Delivery
Experience & Impact
Nov 2025 — Present
Lead Software Engineer
NBCUniversal · Universal Creative and UDX
Led delivery of 10 products from concept into production across 6 business functions. Work spans a full-stack estimating platform, contract lifecycle software, retrieval and evaluation infrastructure, and cost intelligence.
Estimating and takeoff time reduced from 6 hours to 30 minutes · Approximately 900 automated tests backing the platform.
Aug 2024 — Nov 2025
AI Engineer
FOX Sports
Built AI solutions for media post-production: a Slack agent for ML pipeline monitoring, a natural-language-to-SQL assistant, evaluation tooling, and automated metadata extraction.
35% less issue-resolution time with the Slack agent · 40% less ad-hoc reporting effort with the conversational assistant.
04 / Background
About
I'm a software engineer focused on the work around the model: connecting it to the right data, evaluating its behavior, and supporting it in production. Earlier work at Sakar Robotics and a Master's in Robotics at the University of Maryland built my foundation in machine learning, computer vision, and software engineering.
05 / Get in touch
Contact
Let's talk about useful AI systems and the software that makes them work.
ochittar@gmail.com