My Projects
Finbud
An AI-powered personal wealth manager that empowers financial growth with data-driven financial decisions. At Finbud, I architected and deployed an end-to-end Retrieval-Augmented Generation pipeline with Python and Selenium, optimized Qdrant for faster similarity search, and automated Azure VM lifecycles with PowerShell. I also containerized the system with Docker and built a Vue.js chat UI that improved engagement. Beyond the technical side, I learned how to balance performance optimization with cost reduction in a startup setting, and how to communicate complex systems to stakeholders in live demos, shaping the product story for Y Combinator.
British Airways Reviews Dashboard
A full-stack analytics dashboard that processes 3,000+ passenger reviews in real time. I worked on a cross-functional team to build a Next.js frontend with 15+ interactive charts, designed APIs with Node.js/Express to serve data from AWS S3, and streamlined pipelines using Snowflake, EventBridge, and Lambda. I also set up CloudWatch monitoring and CI/CD with GitHub Actions and Jest. This project taught me how to think about systems end-to-end: not just building features, but ensuring reliability (99.9% uptime), scalability, and testability. It also deepened my understanding of how cloud-native pipelines power real-world analytics platforms.
EPL Players Stats
A full-stack web application delivering English Premier League player statistics. I built a Spring Boot backend with REST APIs and PostgreSQL, a React.js frontend with sortable tables and player cards, and automated deployment with Terraform and AWS. I also secured APIs with Spring Security/JWT and built a testing suite with JUnit, Mockito, and Testcontainers. Through this project, I learned the challenges of designing secure and scalable APIs from scratch and how DevOps tools (Terraform, Docker, GitHub Actions) can transform manual setup into repeatable infrastructure. It gave me a deeper appreciation of bridging backend data models with an intuitive frontend experience.
Go Database Internals
A lightweight database engine implemented in Go to explore how databases work under the hood. I built B-tree indexing with node splitting, applied copy-on-write and double-write crash recovery, and optimized page layouts with freelists and offsets. I also benchmarked LSM-trees for write-heavy workloads. This project pushed me to think like a systems engineer: I gained hands-on appreciation for tradeoffs between write speed and read performance, and how low-level memory structures directly shape high-level database reliability.
Career Foundation Hub
A career support platform to help users navigate job applications and development resources. I collaborated with designers to build a responsive Next.js frontend, implemented Firebase authentication with MongoDB storage, and added a Retrieval-Augmented Generation chatbot for career-related queries. What I learned most was how to prioritize UX: making the chatbot intuitive, ensuring authentication felt seamless, and applying feedback loops to iterate on design. It showed me the importance of aligning technical features with user needs.
Forever - E-Commerce Website
A scalable e-commerce web application built with the MERN stack. I built 15+ reusable React components, implemented a persistent shopping cart with React Context, and optimized MongoDB schema design for 10,000+ products. I also styled the UI with Tailwind CSS and semantic HTML for responsiveness. Beyond the technical build, this project helped me understand product thinking: how small frontend details like layout shifts affect user trust, and how schema choices on the backend can shape performance at scale.
