Kunj Shah | Portfolio

LLM Research & Development

How LLMs Are Made

An all-in-one GitHub repo documenting my hands-on journey building and experimenting with LLMs—from GPT, Deepseek, and Kimi architectures to advanced techniques like MoE, MoD, MHLA, and MLA. Includes code, experiments, insights, and resources.

Technical Insights & Documentation

Full-Stack LLM Solutions

Kimi, GPT, Deepseek Architectures

5BooksLM - GPT Model

Designed and trained a 211M parameter GPT-style language model from scratch with 18 transformer blocks and a 512-token context window. Built on 5 different books with full implementation including tokenization, attention mechanisms, training loop, and inference.

211M Parameters

18 Transformer Blocks

512 Token Context Window

Experience

AI Agent Intern, Dreamable Inc.
Collaborating on the development of AI agent solutions that enable developers to build intelligent LLM workflows 2–3x faster using low-code and no-code tools, focusing on agent orchestration, prompt chaining, and modular tool integration to simplify complex AI behavior into developer-friendly components.
- n8n outreach agent adopted by 14 interns
- lead‑quality ↑2.3×
June 2024 to Present
Tech Director, SparkSF
Developed a responsive website for SparksF, an entrepreneurship club at SF State University, to streamline member access to announcements and account management, enhancing engagement through user authentication and real-time updates, along with a chatbot just for SparksF.
Website

Dec 2024 to Present
Intern, Dyna Grow Design Solutions
Built a responsive, user-friendly website using ReactJS and Tailwindcss to enhance Dyna Grow Design Solutions' digital presence, earning a Letter of Recommendation from the founder for delivering lasting impact.
- SEO & perf tweaks drove 2× qualified inquiries
May 2024 to January 2025

Projects

Max - AI Voice Assistant

90% voice accuracy 8 tools Langchain/OpenAI

Developed a voice-activated AI assistant using Langchain, OpenAI, Hugging Face, and SpeechRecognition to automate tasks like web search, YouTube streaming, and emailing, enhancing user experience through hands-free interaction.

Github

FLAN-T5 Stack Overflow Finetuning thumbnail

Llama-3.2-3b FInetuned on OpenHermes

~300k QA Pairs LoRA Finetuning 1.27->0.21 Train loss

An instruction-tuned Llama-3.2-3B base model trained with LoRA on the OpenHermes dataset, reducing eval loss from 6.27 → 2.01. This run transformed the base model into an instruct-capable assistant with only ~0.75% of parameters updated, making it lightweight and deployment-friendly for chatbots and AI agents.

Hugging Face | GitHub

theHelper - AIResearchAssistant

70% faster review 50+ PDFs BERT/BART/FAISS

Built a RAG System to analyze and summarize PDFs using PyPDF2, Hugging Face Transformers (BERT, BART), and FAISS for semantic search, delivering a Streamlit-based app for document processing and context-aware Q&A.

Github

More projects on Github

Technical Skills

Programming Languages

Python

JavaScript

Java

C++

HTML/CSS

SQL

AI Tools & Frameworks

LangChain

LangFlow

n8n

RAG

LLMs

Transformers

OpenAI API

Anthropic API

Hugging Face

MCP Servers

Vector Databases

Prompt Engineering

Machine Learning & Deep Learning

PyTorch

TensorFlow

Scikit-learn

Keras

OpenCV

Pandas

NumPy

Matplotlib

NLP

Computer Vision

Transfer Learning

Model Finetuning

LoRA

Neural Networks

Web Development

Node.js

React.js

Flask

TailwindCSS

Express.js

Database & Development Tools

MongoDB

MySQL

Git

Docker

VertexAI

Microsoft Azure

Hackathons

MCP AWS Agentic Challange

Where: AWS Builder Loft, SF

When: 7/25/2025

Project: Nango Automation

Cal Hacks 11.0

Where: San Francisco, CA

When: October 18, 2024 – October 20, 2024

Project: Workout Web App

SacHacks

Where: Virtual Hackathon

When: March 2, 2025 – March 3, 2025

Project: Web Detective

HackMerced

Where: University of California, Merced

When: March 9, 2025 – March 11, 2025

Project: Web Detective (Updated)

Certificates

Introduction to Generative AI – Google Cloud
TensorFlow Developer – Google
Hugging Face Transformers – Hugging Face

Show all certificates

Programming for Everybody (Getting Started with Python) – University of Michigan
Python Data Structures – University of Michigan
Crash course on Python – Google
Calculus through data and modelling: Series and integration – Johns Hopkins University
Calculus through data and modelling: Techniques of integration – Johns Hopkins University
Calculus through data and modelling: Integration Applications – Johns Hopkins University
Calculus through data and modelling: Vector Calculus – Johns Hopkins University
Introduction to Web Development – UC Davis
Understanding Einstein: Special theory of relativity – Stanford University
Introduction to complex analysis – Wesleyan University
Understanding Basic SQL Syntax – Coursera Project Network
C++ Basics: Selection and Iteration – Codec
Building a Text-based Bank – Coursera Project Network
Create a Supermarket app using Java OOP – Coursera Project Network
Python 101: Develop Your First Python Program – Coursera Project Network
LOR by Duc Ta - CSC215
LOR by Maitra Shah - Internship Certificate

About Me

I am a second-year Computer Science student at San Francisco State University with expertise in AI/ML and Full-stack development. Currently serving as Tech Director at SparkSF, I specialize in Machine Learning, NLP, and MERN stack development. My notable projects include AI-powered applications like 'theHelper' research assistant and 'Max' voice assistant. I've participated in multiple hackathons including Cal Hacks 11.0 and SacHacks, creating innovative solutions like Workout Web App and Web Detective. With strong foundations in Python, JavaScript, and various AI frameworks including Hugging Face Transformers and OpenCV, I combine academic excellence (JEE qualifier) with practical development experience to deliver impactful solutions.