Architected a cloud-native customer support chatbot using AWS Lambda, API Gateway, and S3, reducing inquiry resolution time by 30% through intelligent automation

Designed a RAG pipeline with AWS Textract and PySpark to vectorize 10M+ documents into Pinecone, improving answer accuracy by 23% and cutting irrelevant responses by 40%

Built a production-grade Generative AI Q&A system with FAISS vector search and open-source LLMs, improving factual accuracy by 85% and reducing hallucinations by 70%

Developed high-throughput REST and GraphQL APIs handling millions of concurrent requests with schema validation, rate-limiting, and backward compatibility

Optimized LLM inference pipelines via caching, request batching, and prompt versioning — cutting latency by 20% and cloud costs by 15%

Automated CI/CD pipelines using AWS CodePipeline and Spinnaker with canary deployments and auto-rollbacks, reducing production incidents by 35% across 10M+ daily users

Implemented real-time sentiment analysis with Hugging Face Transformers to classify customer queries and trigger proactive escalation for distressed users

Enforced zero-trust security with JWT, OAuth2, and RBAC across multi-tenant systems handling sensitive user data

Worked as a full-stack engineer building and supporting enterprise web applications using Angular and React on the frontend, with Python (Flask) and Node.js on the backend.

Built AI-powered features using LangChain and Hugging Face, including SQL chatbots and workflow automation tools that improved team productivity.

Helped develop a low-code / no-code platform with drag-and-drop capabilities, enabling non-technical users to create applications and dashboards.

Designed and maintained RESTful APIs within a microservices architecture, focusing on secure and reliable system integration.

Implemented JWT-based authentication and Single Sign-On (SSO) to strengthen security and simplify user access.

Optimized MySQL and MongoDB queries and schemas to improve performance and reliability under higher traffic.

Containerized services using Docker and supported automated deployments through CI/CD pipelines.

Designed enterprise ETL pipelines for high-volume financial data using Java, Spring Boot, and SQL — processing millions of records daily within strict SLA windows

Engineered an AI SQL chatbot using LangChain and Hugging Face NLP, reducing manual query effort by ~40% and improving enterprise data access efficiency

Built high-performance microservices with Spring Boot and gRPC, reducing payment processing latency by 40% and supporting 5,000+ TPS during peak traffic

Fine-tuned open-source LLMs using LoRA/PEFT with Hugging Face Transformers, evaluated via BLEU, ROUGE, and hallucination benchmarks — outperforming baseline models

Optimized PostgreSQL, MySQL, and MongoDB through indexing and query refactoring, improving performance by 30%+ on production systems

Designed React.js + Redux frontend component libraries, accelerating UI development by 30% and improving Core Web Vitals scores by 15%

Containerized services with Docker and deployed on AWS EC2/Lambda with Jenkins CI/CD, improving release velocity and reducing deployment failures

Integrated Azure AI Document Intelligence with custom NLP/OCR pipelines to automate high-volume form processing, achieving 91% extraction accuracy and reducing manual data entry costs by 45%.

Developed and deployed full-stack e-commerce platforms using Angular 12, Node.js, and MongoDB, increasing conversion rates by 22% and reducing cart abandonment by 15% through UX and performance optimizations.

Designed intuitive UI/UX workflows with Figma and Angular Material, leveraging A/B testing to boost engagement by 30% and cut support tickets by 40%.

Architected secure, scalable REST APIs handling thousands of concurrent requests with versioning, schema validation, authentication, and rate limiting to ensure reliability and backward compatibility.

Built CI/CD pipelines with Jenkins, Azure DevOps, and GitHub Actions, implementing Blue-Green deployments for zero-downtime releases and improved release stability.

Integrated third-party APIs across frontend and backend systems, enhancing interoperability and improving end-to-end data flow efficiency by 25%.

Developed responsive web and mobile applications using Angular and Ionic, ensuring seamless cross-platform performance and consistent user experiences.

Engineered interactive, data-driven UI components with Konva.js, enabling real-time drawing, annotations, and dynamic labeling to significantly enhance usability.

Collaborated with backend, QA, and product teams to deliver reliable, user-focused interfaces, improving overall application efficiency by 25%.

Optimized frontend performance through debugging, refactoring, and efficient state management, resolving critical UI bottlenecks and improving responsiveness.

Supported testing, UAT, and release readiness while maintaining code quality through Git-based workflows and automated CI/CD processes, enabling stable production deployments and continuous product enhancements.