AI/ML Data Engineer

Developed, trained, tested, and deployed machine learning models for business applications. Built scalable data pipelines processing more than 5 million records daily, reducing manual data preparation by 60%.

Implemented AI/ML solutions using Python, Scikit-learn, TensorFlow, and PyTorch, with data quality validation and monitoring. Processed structured and unstructured data for feature engineering, model development, and RAG pipelines using Pinecone and Chroma.

Productionized data flows for agentic systems through four production microservices handling over 10,000 daily requests with 99.9% uptime. Optimized PostgreSQL and MongoDB queries, reducing API response time from 800 ms to 210 ms while supporting threefold traffic growth.

Contributed to technical documentation, design discussions, code reviews, CI/CD workflows with GitHub Actions, and Generative AI implementations including prompt engineering and hallucination control.

Engineered ETL systems processing more than 2 million racing records for real-time feature engineering and predictive analytics using Python and AWS Lambda.

Developed ML models with LangChain-powered agents and OpenAI embeddings. Implemented semantic search and optimized orchestration to reduce inference latency by 45%.

Built Pinecone-backed vector search for dynamic calculations, trend analysis, and intelligent product recommendations. Deployed services through AWS Lambda and API Gateway with auto-scaling, achieving 99.8% uptime during peak traffic.

Developed scalable microservices serving more than 8,000 learners, using vector databases and OpenAI APIs to provide semantic search and enhanced information retrieval.

Designed feature stores and data pipelines for personalized content recommendations, increasing average session duration by 30%. Integrated RAG-based question-answering systems that reduced support tickets by 40%.

Containerized services with Docker and automated zero-downtime deployments through CI/CD pipelines using GitHub Actions.

Built data pipelines for an intelligent e-commerce platform handling more than 500 daily transactions, including JWT authentication and intelligent product search.

Implemented a recommendation system using OpenAI embeddings and collaborative filtering, improving conversion rates by 15%. Designed a Pinecone-based product similarity engine that reduced search response time from 900 ms to 180 ms.

Deployed microservices on AWS EC2 with auto-scaling and Redis caching to handle flash-sale traffic spikes and maintain production reliability.

Built Python and NLP-based data pipelines with secure validation and debugging workflows for a regulated environment.

Improved platform reliability by 25% through automated log analysis and real-time anomaly detection across game sessions. Integrated a fine-tuned BERT model for intent classification, improving accuracy from 78% to 93%.

Created automated debugging pipelines and session replay analytics, reducing user-reported issues by 30%.

Completed a Master of Computer Applications program in Bangalore, India.