Software Engineer at Netgear

Architected a scalable enterprise banking application using React, Java, Spring Boot, and MongoDB, improving user engagement by 40% via streamlined account management and real-time transaction tracking. Designed and deployed RESTful APIs following OOD principles supporting 6,000+ daily transactions with 99.9% availability. Implemented optimized MongoDB schemas and indexing for 20,000+ users, reducing schema design time by 30% using GitHub Copilot. Built CI/CD pipelines with Jenkins, Docker, and GitHub while integrating Apache Kafka for real-time event streaming, cutting deployment time by 50%. Developed a RAG-based LLM pipeline using ChromaDB and Google Gemini with 150+ daily context-aware lookups at sub-3-second latency, and deployed ML pipelines on AWS Lambda/S3 for predictive analytics.

Graduate Research Assistant at UNC Charlotte

Engineered a multi-agent AI system simulating human decision-making for behavioral modeling, improving response consistency by 35% over baseline configurations (measured by human evaluators). Generated 10,000+ synthetic behavioral samples using generative AI to train an Emotion Engine, improving human-centric interaction accuracy across diverse contexts. Built automated web scraping and labeling pipelines to collect and clean 12,000+ multimodal (text/image) examples for fine-tuning Transformer-based emotion recognition models. Fine-tuned open-source LLMs on emotion-labeled datasets using instruction tuning and preference-based feedback for improved detection accuracy across text and image inputs. Established scalable AWS S3 data storage and versioning workflows, reducing data preparation time for fine-tuning by 40%.

Software Developer at RSM US LLP

Enhanced ReactJS/TypeScript frontend features (debounced inputs, micro frontends, real-time validations), improving data input efficiency by 50% for tax filing workflows. Optimized frontend performance using lazy loading, code-splitting, and server-side rendering in Redux with RTK Query, reducing page load times by 25% across dashboards. Built backend services using Java, Spring Boot, and MySQL to handle 50,000+ tax records annually while reducing downtime by 45% via robust error handling. Developed distributed event-driven microservices with Spring Boot and Azure AI Document Intelligence to extract structured JSON from 3,000+ PDFs/images with 95% accuracy. Streamlined ETL pipelines with Azure Data Factory to convert raw client data into clean MySQL datasets, reducing processing time by 30%.

Software Engineer at Mindtree

Developed full-stack features for enterprise clients using Java, Spring Boot, React, and MySQL, delivering 15+ production-ready modules processing 100,000+ daily API requests. Implemented RESTful Spring MVC microservices and integrated Kafka for asynchronous event handling, reducing inter-service latency by 35%. Automated CI/CD pipelines using Jenkins and Docker, cutting manual deployment effort by 60% and enabling zero-downtime releases across staging and production. Collaborated on migrating on-premise applications to AWS (EC2, S3, RDS), achieving 99.95% uptime and reducing infrastructure costs by 20%. Wrote unit/integration tests (JUnit, Mockito) achieving 90% code coverage and mentored two junior developers in Agile practices.