Ten years of streaming pipelines, serverless orchestration, and enterprise data platforms — now channeled into multi-agent LLM systems.
About
Senior engineer with 10+ years building large-scale, cloud-native data and application systems — real-time Kafka streaming pipelines, AWS serverless orchestration, and enterprise-scale ETL, mostly in regulated healthcare. For the past year I've been applying that foundation to AI: multi-agent orchestration with LangGraph, retrieval-augmented generation over vector stores, and tool-calling agents. My focus now is productionizing LLM systems — deployment, evaluation, and observability.
Selected work
Grouped by what they are. The agentic systems are where I'm heading; the data & ML work is the foundation it's built on.
// Agentic & LLM systems
A multi-agent LangGraph system that helps rural patients find providers, understand symptoms, and prepare for visits. Supervisor-pattern agent graph (triage, resource finder, insurance checker, visit prep), RAG over a Qdrant vector store, tool-calling agents for geocoding and provider registries, structured output, and human-in-the-loop review.
An LLM-powered reader that ingests insurance benefit documents and returns structured, queryable data instead of dense PDFs.
Currently building through the Agentic AI 3.0 track — MCP servers, agent evaluation, and multi-framework orchestration. New projects land here as they ship.
// Data & ML
Supervised machine learning predicting ADHD in women from a human brain-mapping dataset, built for a 2025 Kaggle datathon.
A batch data-engineering pipeline analyzing Netflix movie data, with rating- and vote-based recommendation logic and genre rollups.
In progress
// Agentic AI 3.0 — Krish Naik Academy. These graduate into projects above as I build with them.
Experience
Toolkit
Things I've built with in production or in projects.
LangGraph · LangChain · Retrieval-Augmented Generation (RAG) · Qdrant (vector DB) · tool-calling / function calling · structured output · multi-agent orchestration
Python · scikit-learn · pandas · NumPy · Jupyter · supervised learning
AWS (Lambda · Step Functions · SQS · SNS · API Gateway · S3 · Glue · Kinesis · DynamoDB · RDS · CloudFormation · EC2 · IAM · KMS) · Apache Kafka · PySpark · ETL / ELT · Terraform · GCP
Java · Python · Node.js · Go · SQL
Spring Boot · microservices · REST & SOAP APIs · OpenAPI/Swagger · Docker · Kubernetes · OpenShift · PostgreSQL · MongoDB · Oracle · SQL Server
CI/CD (Jenkins · GitLab) · Maven · Agile / Scrum · TestNG · Cucumber BDD · Splunk