Rajeev Nandan

AI & Full-Stack Engineer

Generative AI Intern @ IIT Kanpur · 3× Hugging Face deployments · GPA 9.24

Looking for roles in AI/ML Engineering or Full-Stack Development

Download Resume View Projects

Open to opportunities

AI/ML Engineer

B.Tech CSE · GITAM University, Visakhapatnam

AI Engineering with Product & Deployment Impact

Delivered 3+ production Hugging Face deployments with real user-facing value.
Built scalable Next.js + TypeScript full-stack systems for modern AI experiences.
Optimized ML inference and backend workflows to improve performance and reliability.

Scroll to review high-impact AI projects, architecture, and execution details.

About Me

Bridging core computer science theory with scalable, production-ready AI pipelines.

As a Computer Science student at GITAM, I do not just train ML models—I optimize them for high-performance serving. I specialize in Generative AI systems, focusing on Stable Diffusion tuning, multi-document RAG pipelines, and sub-40ms real-time vision pipelines.

Research

IIT Kanpur Intern

Stable Diffusion & LLM baseline tuning

Deployment

3+ Live ML Spaces

Production serving deployed on Hugging Face

Education

GITAM University, Visakhapatnam

B.Tech Computer Science & Engineering

2023–Present

0.00/10

Cumulative CGPA • Latest Sem GPA: 9.73/10

Windows PowerShell

⎯⧠✕

Windows PowerShell
Copyright (C) Microsoft Corporation. All rights reserved.

Try the new cross-platform PowerShell https://aka.ms/pscore6

Try command:

Hugging Face Live Sandbox

Insight GPT Space

CPU: 14%

Stable Diffusion Space

CPU: 8%

YOLOv8 Edge Feed

CPU: 21%

Applied CS Theory

Click on each category to explore how textbook computer science principles are directly integrated into high-performance ML pipelines.

Artificial Intelligence & NLP

Click to inspect applied project metrics

Applying deep neural networks, language modeling, and image synthesis algorithms.

Production Application

Configured Stable Diffusion noise scheduling parameters in PyTorch to cut FID scores by 15%. Engineered scalable LangChain RAG architectures using semantic chunking.

Stable Diffusion TuningRAG Vector PipelinesGPT Prompt TaxonomyYOLOv8 Edge ML

Systems & Algorithms

Click to inspect applied project metrics

Network & Security

Click to inspect applied project metrics

Python • PyTorch • Hugging Face • Transformers • RAG • Vector Databases • Qdrant • FAISS • LLM Fine-Tuning • Stable Diffusion • C++ • TypeScript • SQL • System DesignPython • PyTorch • Hugging Face • Transformers • RAG • Vector Databases • Qdrant • FAISS • LLM Fine-Tuning • Stable Diffusion • C++ • TypeScript • SQL • System DesignPython • PyTorch • Hugging Face • Transformers • RAG • Vector Databases • Qdrant • FAISS • LLM Fine-Tuning • Stable Diffusion • C++ • TypeScript • SQL • System DesignPython • PyTorch • Hugging Face • Transformers • RAG • Vector Databases • Qdrant • FAISS • LLM Fine-Tuning • Stable Diffusion • C++ • TypeScript • SQL • System Design

Next.js • React • Node.js • FastAPI • REST APIs • WebSockets • Docker • AWS • Git • CI/CD • OAuth 2.0 • Secure API Development • Data Encryption • High-Throughput servingNext.js • React • Node.js • FastAPI • REST APIs • WebSockets • Docker • AWS • Git • CI/CD • OAuth 2.0 • Secure API Development • Data Encryption • High-Throughput servingNext.js • React • Node.js • FastAPI • REST APIs • WebSockets • Docker • AWS • Git • CI/CD • OAuth 2.0 • Secure API Development • Data Encryption • High-Throughput servingNext.js • React • Node.js • FastAPI • REST APIs • WebSockets • Docker • AWS • Git • CI/CD • OAuth 2.0 • Secure API Development • Data Encryption • High-Throughput serving

RAG

Vector Embeddings & Semantic Search Pipelines

Diffusion Models

PyTorch & Hugging Face Image Synthesis

FAISS

High-Performance Dense Vector Similarity Search

Gradio

Interactive Web Interfaces for Model Demos

Next.js & TypeScript

Type-safe React architectures, server component state, and premium performance optimization.

Cybersecurity Systems

Security log auditing, intrusion detection, and threat analysis for enterprise operations.

React

Component architecture & dynamic user flows

Node.js

Asynchronous Backend Pipelines & REST APIs

Projects.

A collection of work exploring the intersection of AI, web development, and data science.

Insight GPT

Code

Next.js, Python, FAISS, PyTorch

Edge Object Detector

Code

React, DETR, PyTorch, Quantization

Facial Emotion Classifier

Code

Python, PyTorch, OpenCV, CNN

Latent Diffusion Generator

Code

Python, PyTorch, Hugging Face, Gradio

3D Dark-Ambient Showcase

Code

Next.js, Three.js, GSAP, Tailwind

Customer Segment Analytics

Code

Python, Pandas, Scikit-learn, Jupyter

View

Experience

Professional experience optimizing high-throughput AI systems and delivering research value.

Generative AI Intern

Indian Institute of Technology, Kanpur

June 2025 – August 2025

-15% FID Score-22% Inference Latency+18% Quality Lift

Engineered Stable Diffusion fine-tuning pipelines in PyTorch (DDPM/DDIM), cutting FID score by 15% through custom noise scheduling on domain-specific image datasets
Slashed inference latency by 22% by implementing dynamic batching and FP16 mixed-precision inference across the full diffusion model serving stack
Built a model evaluation harness benchmarking GPT-2, LLaMA, and Mistral across 3 prompt taxonomies — surfacing the highest-performing architecture (18% quality lift) adopted as the team's standard baseline

Verify Certificate→

Independent AI Engineer

Open Source · Hugging Face Spaces

Jan 2025 – Present

Insight GPT (RAG)Stable Diffusion UIYOLOv8 Edge Feed

Shipped Insight GPT — a production RAG pipeline (LangChain + FAISS + OpenAI) with live inference on Hugging Face Spaces, handling multi-document Q&A with semantic chunking and source attribution
Built and deployed a Stable Diffusion UI with custom LoRA fine-tuning pipeline for domain-specific image synthesis, enabling style-transfer at inference time without full retraining
Deployed a real-time YOLOv8 object detection system with webcam inference pipeline achieving sub-40ms end-to-end latency, demonstrating production edge-AI readiness

View Hugging Face Profile→

View All Certifications