Research, development, and innovation projects
Here are some of the projects I've worked on, ranging from research and hardware development to software applications and startup ventures. Each project represents my passion for innovation and problem-solving.
Built a simulation framework integrating Verilator with CUDA to accelerate Verilog testbench execution. Leveraged GPT-based models to suggest targeted test inputs for RTL verification.
Designed software to automate AASHTO-based isometric bearing design calculations for bridges. Architected input-processing-output flow with integration to Mathcad, Excel, and Leap via APIs.
Developed cross-platform mobile app for booking barbers and simulating haircuts via AI face scan. Implemented head shape detection and 3D preview for personalized recommendations.
Created GPT-powered extension to help students by parsing web content and offering inline guidance. Implemented features like "highlight-to-ask," full-page context scanning, and page awareness.
Built mobile app to analyze body composition using mirror selfies and input stats. Generated AI-based fitness goals, diet plans, and chatbot feedback using FastAPI backend.
Created ML pipeline using NewsAPI + yfinance to predict S&P 500 movements from financial headlines. Trained logistic regression and SGD models on TF-IDF vectors of news content.
Built an interactive visualizer that converts hand-drawn shapes into Fourier series. Intended to help students explore signal processing concepts and complex decomposition.
Developed a GPU-powered framework for accelerating RTL simulation and test generation using CUDA and ML.
Contributed to LabGym, an AI-based tool for automated behavioral analysis in neuroscience experiments.
Designed and implemented a 5-band real-time audio equalizer on an FPGA using Verilog.
Implemented a comprehensive heap allocator for x86-64 ISA, optimizing memory usage with implicit and explicit free lists.
Designed and implemented backend services using Python and Flask, including database schema design and RESTful API endpoints.
Co-founded and built an AI-driven iOS app that learns from users' lecture notes, videos, and audio to generate personalized learning plans.
Engineered a motorized couch (2.5M TikTok views) and built an EEG-controlled toy car using Adafruit microcontroller.
My current research at Stanford's Robust Systems Group focuses on building AI-augmented verification systems to analyze and simulate RTL-level designs using LLMs and symbolic execution.