Available for internships & AI roles

Rahul Gupta

AI / ML Engineer in training. MCA student at KNIT Sultanpur building end-to-end AI systems — from ML pipelines to GenAI applications and RAG chatbots.

View Projects → GitHub ↗

82%

Titanic Accuracy

99.1%

MNIST CNN

0.81

Housing R²

Projects Built

// About Me

Building Real AI Systems

I am a first-year MCA student at Kamla Nehru Institute of Technology, Sultanpur, affiliated to AKTU Lucknow. My background in B.Sc Mathematics (Honours) gives me a strong foundation in the statistics and linear algebra behind ML algorithms.

I am not just learning theory — I build real projects with measurable results. Currently in Phase 4 of my AI Engineer roadmap, working on GenAI, LLM APIs, embeddings, vector databases, RAG pipelines, and LangChain.

My target: join an AI-first startup or product company as an AI/ML Engineer after completing my roadmap. Open to internships immediately.

// Technical Skills

What I Know

Core Language

Python

Functions OOP Comprehensions File Handling Lambda

Data Science

Data Libraries

NumPy Pandas Matplotlib EDA Scikit-learn

ML Algorithms

Machine Learning

Random Forest Logistic Reg Decision Tree KNN Model Evaluation

Neural Networks

Deep Learning

PyTorch CNNs Backpropagation Training Loop Adam / BCELoss

In Progress

GenAI & LLMs

Prompt Engineering LLM APIs Embeddings ChromaDB RAG LangChain

Tools

Developer Tools

Git GitHub Google Colab VS Code SQL basics

// Projects

What I've Built

01 — Featured

RAG Study Assistant

Retrieval-augmented Q&A system. Uses ChromaDB as vector store, Sentence Transformers for embeddings, and Llama 3.3 via Groq API. Answers only from uploaded documents — says "I don't have that in my notes" when the answer isn't in the source material. Deployed live on HuggingFace Spaces.

Python ChromaDB Groq API Streamlit

Live Demo → View Code →

RAG FastAPI Service

Same retrieval pipeline exposed as a REST API instead of a UI. FastAPI + Pydantic, containerized with Docker so it can be called from any client.

FastAPI Docker Pydantic

REST API View →

NayePankh Chatbot

RAG-based chatbot built as a hiring task for NayePankh Foundation, a UP-registered NGO. Answers questions about the foundation using retrieved context, not hardcoded FAQs.

LangChain ChromaDB Groq API

Live Demo → View →

ML Fundamentals — Titanic & Housing

Two foundational ML projects. Titanic: classification with Random Forest. California Housing: regression comparing Linear Regression vs Random Forest.

Scikit-learn Pandas

82% acc · R²=0.81 View →

MNIST Digit Recognition — CNN

Convolutional Neural Network in PyTorch, trained on 70,000 handwritten digit images. Conv2d, MaxPooling, Adam optimizer.

PyTorch CNN

99.10% Accuracy View →