Principal Engineer at Optum/UnitedHealth Group's Chief Data Office, transitioning 24+ years of engineering experience into AI/ML and Agentic AI.
- 🤖 Building ADPA ("Optimus Prime") — an Agentic-AI platform that turns a data contract into a production-ready data product on Azure Databricks/Snowflake in hours instead of weeks (114K+ lines, 2,900+ tests, placed 3rd in company-wide in an internal AI innovation event)
- 🧠 Recently completed the MIT Professional Education Applied Data Science Program (Leveraging AI for Effective Decision-Making) and a Generative AI certification
- 🏗️ Background in Data Mesh architecture, Azure Databricks, Apache Kafka/Spark, and large-scale Hadoop/Big Data systems
- ⚙️ Founding engineer of a Go-based streaming filter library with a custom filter DSL (nested boolean/array logic, EBaaS caching, CLI tooling) — still powering 35+ production applications today
- 📜 Named inventor on a submitted patent for a rule-based runtime transformation layer for FHIR healthcare data
- 🎓 Diamond Bravo Award recipient for mentoring an early-career team through a Care Coordination & Referral Tracking microservice capstone in Optum's Technology Development Program (TDP) Bootcamp
- 🚀 Currently exploring: LangGraph, multi-agent systems, RAG architectures
- 📊 Built and benchmarked recommender systems (SVD, KNN, matrix factorization) on the Amazon Electronics dataset as part of the MIT program, reaching RMSE 0.8808 and F1@10 0.869 on tuned models
Core stack: Python · Scala · Go · TypeScript · LangGraph · Apache Spark · Kafka · Databricks · Snowflake · Apache Iceberg · Terraform · Kubernetes · HL7 FHIR · scikit-learn · TensorFlow · pandas · NumPy
📫 Reach me: LinkedIn · thiruselvaa@gmail.com