I'm a Generative AI Engineer with 5+ years of experience building and shipping production systems. Most of my work sits at the intersection of LLM apps and backend engineering: search and retrieval over real business data, agent workflows for multi-step requests, and APIs that stay stable under real usage. I've worked across supply chain, banking, healthcare, energy, and manufacturing, and I'm comfortable taking a feature from data prep to deployment.
About
● updated recentlyI build GenAI features that people can rely on day-to-day. That usually means turning messy documents and operational data into something searchable, then wiring it into a clean workflow that returns accurate, grounded answers. When the use case needs more than one step, I add agent-style orchestration to handle routing, checks, and follow-ups.
I care a lot about the "boring but important" parts: retrieval quality, stable prompts, output validation, access control, and good logging/monitoring. I also keep deployments simple with CI/CD and containerized releases, so shipping changes doesn't turn into a fire drill.
Experience
● updated recently
Built and deployed GenAI solutions for supply-chain planning and forecasting using Azure OpenAI, LangChain, LangGraph, and FastAPI. Implemented RAG over SOPs, planning documents, and operational data using Azure Cognitive Search and embedding models, and used agent workflows for multi-step scenarios that needed routing and validation. Developed backend services in Python and Go, built ingestion pipelines with Azure Data Factory and Databricks, and improved quality through prompt tuning, response checks, and regression testing. Owned production readiness with security controls, CI/CD, and monitoring using AKS, Docker, GitHub Actions, Prometheus, and Grafana.
Worked on production AI and GenAI systems in a regulated banking environment. Built retrieval-based solutions over policies and historical case data using OpenSearch and embedding models, and designed agent workflows for compliance reasoning and validation. Developed document intelligence pipelines with Textract, built scalable inference services using FastAPI and Docker, and supported deployments on AWS using SageMaker, EKS, and Terraform. Focused heavily on governance, auditability, monitoring, and reliability to meet regulatory requirements.
Delivered AI and GenAI features for smart-energy analytics, combining LLM-based insights with predictive modeling and large-scale IoT data pipelines. Built agent workflows using LangChain and CrewAI, implemented retrieval-based search over grid and maintenance data using OpenSearch and FAISS, and developed LLM-enabled applications using Amazon Bedrock and Hugging Face models. Built forecasting and anomaly detection models with SageMaker and TensorFlow, and supported high-volume ingestion using IoT Core, Kinesis, Glue, and Spark on EMR. Deployed services through cloud-native patterns with FastAPI, containerization, and automated delivery.
Built healthcare applications and backend services using Python with Flask and Django, focused on reliable APIs for clinical workflows and analytics. Developed and deployed ML and deep learning models for NLP, computer vision, and time-series use cases, and set up reproducible training and deployment using Kubeflow on GKE. Used Vertex AI for experimentation and deployment, and built data pipelines with BigQuery and Dataflow to support reporting and segmentation. Focused on clean service design, stable integrations, and production support in an Agile environment.
Built backend and platform services supporting enterprise manufacturing workflows. Developed Python microservices using Flask, Django, and FastAPI, and deployed them using Azure Functions and AKS. Implemented API gateways with authentication, routing, and rate limiting, and built event-driven pipelines using Kafka and Azure messaging services. Worked extensively on CI/CD, infrastructure automation with Terraform and Ansible, and improving performance and reliability of production systems.
Worked on full-stack and backend systems for data-driven enterprise applications. Built Django-based web applications, internal tools, and admin workflows, and supported data processing using Python and Spark. Contributed to CI pipelines using Jenkins and Git, handled Linux-based deployments, and worked with AWS services for application hosting and data workflows. Gained strong experience in building maintainable systems and working in Agile delivery environments.
Projects
● updated recentlyA production-grade GenAI-powered conversational agent for exploring IMDB's Top 1000 movies. The system features intelligent query routing that automatically classifies queries as structured (SQL), semantic (vector search), or hybrid, then executes the best strategy and synthesizes a grounded response.
A local Retrieval-Augmented Generation (RAG) system that lets users upload PDF documents and ask natural-language questions using a locally running LLM via Ollama — no paid APIs required. Built with a production-style FastAPI backend, FAISS vector store, and a clean CI pipeline.
Education
● updated recently