👋

Hello, I'm Ameya.

About me

I’m a student at Stanford studying Computer Science with a concentration in Artificial Intelligence and a minor in Management Science and Engineering. I've also been an undergrad researcher at Stanford Artificial Intelligence Lab — working on LLMs.

At Stanford, I'm a VP with Cardinal Ventures and Chief Financial Officer of TreeHacks.

featured projects

TestNinja

Autonomous agent platform for robust and context-aware test generation.

Next.js
Llama 3.1 (Baseten)
Python
Flask
ESPrima

Dave

Devin-esque SWE agent that interacts with local development tools

React
Next.js
Docker
LangChain
Python

Sonoverse

Protecting small artists' work through a decentralized on-chain ML platform, integrating automatic DMCA claims.

Cohere
Ethers
Next.js
Pytorch
Python
Caldera
Solidity

Mixer

The best platform to organize and plan college events.

GCP
React
React Native
Tailwind
Node.js

Quasar

An intelligent shopping platform that utilizes AR and ML to find the product you need. Crypto friendly and NCR-services compatible.

React
CryptoJS
Echo3D
TensorFlow

My Skin

Your personal skin monitoring app - a simple AI tool to assist and help you explore and monitor changes on your skin and moles.

Swift
Keras
Python
Tensorflow

Research

Enhancing LLM General Reasoning with ITPs

Stanford STAIR Lab (SAIL)

This study introduces a method to improve Large Language Models' (LLMs) mathematical reasoning and autoformalization capabilities by incorporating formal proofs from Interactive Theorem Provers (ITPs) like Lean, Coq, and Isabelle. Initial results show modest improvements in LLMs' ability to generate informal proofs, suggesting the potential of formal proofs to advance LLMs' mathematical reasoning.

Paper Code

Medical NER and RE from Clinical Notes

Stanford CS Department

This project develops a system for medical named entity recognition (NER) and relation extraction (RE) from clinical notes, aiming to identify medical concepts and their relationships. The goal is to automate the extraction of a patient’s treatment history and effective medications, providing comprehensive profiles for better clinical decision-making

Paper

Digital Threats Against Democracy

Georgia Tech T+ID Lab

NLP research on Twitter social networks and how communities cross-interact; and HCI augmented research on how human trackers can monitor multiple social media feeds in semi real-time.

Project Website