Projects

An LLM-powered mock interview simulator. Pick an interview type, a scenario, and a duration; Claude plays the interviewer, asks phase-aware questions, and scores your performance at the end.

AI web app for due-diligence financial document analysis. Integrates a Langchain RAG pipeline and Pinecone vector database for robust document understanding.

An AI-powered next-action engine that helps you break through decision paralysis and stay in flow. Surfaces your most important task based on context, energy, and deadline signals.

A Chrome extension that coaches you through DSA problems in real time. Detects which LeetCode problem you're on, tracks your thinking, and gives Socratic hints — never just the answer.

CRNN models for handwriting text recognition trained on GCP. Achieved 12% word accuracy improvement using word beam search and CTC loss.

A full operating-system kernel built in C for Stanford's OS course (CS 140). Implements a priority-based thread scheduler, virtual memory, 14 system calls, and an indexed-inode file system.