Current role focusing on biomedical QA with retrieval-augmented generation. Improved Q&A accuracy by 41% and Recall@10 from 0.58 to 0.91 by engineering a RAG agent using LangChain, FAISS vector DB, BioBERT embeddings, and adaptive chunking across 50K+ PubMed abstracts. Accelerated feature delivery from 8 to 3 days by integrating Codex, Cursor, and Claude AI tools for scaffolding, refactoring, and test generation across 15+ production features with a 94% code review pass rate.

Software Engineer Intern at NYS Office of Information Technology Services

June 1, 2024 - August 1, 2024

Delivered a 98% WCAG 2.1 AA-compliant, accessible React portal enabling 140K+ e-filings by developing a real-time form with pre-validation and prior-year auto-fill using JavaScript, Node.js, PostgreSQL, and RESTful APIs.

Software Engineer II at Zemoso Technologies Pvt. Ltd.

July 1, 2022 - July 1, 2023

Led an LLM-based fraud detection system reducing bot/spam accounts by 70% with 93% detection precision using FastAPI, React, JavaScript, AWS Lambda, and PostgreSQL. Built a distributed event-driven logistics backend, cutting API latency from 320ms to 95ms and achieving 99.9% uptime using Node.js, Apache Kafka, AWS Lambda, and PostgreSQL. Boosted CI/CD pipeline speed 3x by containerizing Node.js microservices with Docker and Kubernetes, enabling parallel GitHub Actions workflows. Architected a JSON Schema-driven React form system with geo-location, signature, and image inputs, reducing development time by 60% for 5,000+ monthly technicians via React Hook Form, Redux, and reusable components. Built offline-first capabilities for an iOS/iPad React Native app with 50ms reads and 200ms writes via Realm DB, custom caching queues, OOP-driven backend synchronization, and conflict detection/resolution.

Associate Software Engineer at Zemoso Technologies Pvt. Ltd.

April 1, 2021 - June 1, 2022

Led the migration of a legacy monolith to 12 distributed event-driven microservices, reducing infrastructure costs by 70% and achieving 99.99% uptime using Apache Kafka, Terraform, and AWS Lambda. Engineered an offline-first React Native iOS/iPad app with 11 modules for 5K+ DAU, reducing launch time from 5s to 2s by lazy loading, Redux optimization, and Node.js, Apollo GraphQL on AWS via GitHub Actions CI/CD. Automated a React testing pipeline achieving 98% test coverage and saving 3 hours per sprint by implementing TDD across 250+ test cases with Jest and Detox, integrated into GitHub Actions CI/CD pipelines.