Over the past few years built some awesome projects surrounding few topics. I also interned at Tarum (a Generative AI startup), where I worked across different microservices (in Golang), shipped new features, identified and resolved payment service flaws, and debugged Kubernetes pod issues. I am also part of the Robotics Club, working towards competing in the RoboCup SSL Simulator.
In my free time I also like to re-implement papers from scratch in different programming languages.
Final Year Project: automated tool to detect violations in website privacy policies and cookie banners against GDPR & PDPA. Automating cookie consent management and also detecting dark patterns in cookie banners.
Built an agentic RAG system using 35+ domain research papers with vector embeddings, retrieval pipelines, and LLM orchestration for multi-step reasoning and tool use. Implemented end-to-end monitoring with Prometheus and Grafana dashboards tracking context precision, answer relevance, faithfulness metrics, and system latency.
A system designed to handle background tasks concurrently to improve user experience. Instead of blocking API calls for tasks like sending welcome emails on login, jobs are added to a WorkQueue and executed in the background via Redis workers.
Enables real-time chatting and file sharing on your local network, using mDNS and WebSocket for seamless, server-free communication.
A Go HTTP server demonstrating idempotent request handling using Redis caching and PostgreSQL persistence. This is behind infrastructure of every payment system, when user clicks twice, request is handled only one time.
Final Year Project: automated tool to detect violations in website privacy policies and cookie banners against GDPR & PDPA. Automating cookie consent management and also detecting dark patterns in cookie banners.
Built an agentic RAG system using 35+ domain research papers with vector embeddings, retrieval pipelines, and LLM orchestration for multi-step reasoning and tool use. Implemented end-to-end monitoring with Prometheus and Grafana dashboards tracking context precision, answer relevance, faithfulness metrics, and system latency.
A system designed to handle background tasks concurrently to improve user experience. Instead of blocking API calls for tasks like sending welcome emails on login, jobs are added to a WorkQueue and executed in the background via Redis workers.
Enables real-time chatting and file sharing on your local network, using mDNS and WebSocket for seamless, server-free communication.
A Go HTTP server demonstrating idempotent request handling using Redis caching and PostgreSQL persistence. This is behind infrastructure of every payment system, when user clicks twice, request is handled only one time.
Final Year Project: automated tool to detect violations in website privacy policies and cookie banners against GDPR & PDPA. Automating cookie consent management and also detecting dark patterns in cookie banners.
Built an agentic RAG system using 35+ domain research papers with vector embeddings, retrieval pipelines, and LLM orchestration for multi-step reasoning and tool use. Implemented end-to-end monitoring with Prometheus and Grafana dashboards tracking context precision, answer relevance, faithfulness metrics, and system latency.
A system designed to handle background tasks concurrently to improve user experience. Instead of blocking API calls for tasks like sending welcome emails on login, jobs are added to a WorkQueue and executed in the background via Redis workers.
Enables real-time chatting and file sharing on your local network, using mDNS and WebSocket for seamless, server-free communication.
A Go HTTP server demonstrating idempotent request handling using Redis caching and PostgreSQL persistence. This is behind infrastructure of every payment system, when user clicks twice, request is handled only one time.