• Constructed a full-stack RAG system using LangChain, OpenAI API, and local Llama models, enabling users to upload and query summarized book content in a vector database.• Built a responsive frontend using React, Next.js, Shadcn and DaisyUI, creating an intuitive interface for seamless access to the book summary system.• Developed a RESTful backend API with FastAPI, integrating PostgreSQL for user data management and applying Redis as a caching layer for user sessions, reducing server load by 30%.• Containerized the application with Docker and deployed it on AWS EKS with Kubernetes for scalability and reliability, managing infrastructure settings with Terraform.• Set up CI/CD pipelines with Jenkins, incorporating unit tests and end-to-end tests to streamline continuous deployment and reduce manual tasks.

• Contributed to the development and tuning of a model designed to detect unwanted elements, such as glasses and necklaces, in user-submitted portraits.• Utilized ONNX format to convert the model from PyTorch to TensorFlow.js, enabling it to run on a React-based frontend interface.