Naveenkumar Narsozhan

MS in Computer Science @ USC | B.Tech @ NIT Puducherry

Turning data into intelligent systems | AI, NLP, & Deep Learning

About Me

I am a Computer Science graduate student with a strong interest in Artificial Intelligence, Machine Learning, and Natural Language Processing. I enjoy designing intelligent systems that transform data into meaningful insights and scalable solutions. With hands-on experience across machine learning, deep learning, and data-driven application development, I focus on building solutions that are practical, reproducible, and impactful.

Currently pursuing my Master’s in Computer Science in the United States, I am actively seeking opportunities where I can apply my technical skills to solve real-world problems and contribute to innovative engineering teams.

Naveenkumar Narsozhan

Technical Skills

Languages

PythonCC++SQLJavaScriptTypeScriptBash

Frameworks & AI

PyTorchTensorFlowHuggingFaceYOLOLangChainLlamaIndexFastAPIScikit-learnOpenCV

Domains

ML/DLNLPComputer VisionAgentic AIRAGLLMsPrompt EngineeringQuery Optimization

Tools & Infra

AzureDockerGitPostgreSQLMongoDBChromaDBVercelSupabaseLinuxCI/CD

Education

University of Southern California

Masters in Computer Science

CGPA: 3.82 / 4.0

Aug 2025 – May 2027

Coursework

Analysis of Algorithms Information Retrieval Deep Learning Machine Learning

National Institute of Technology Puducherry

B.Tech in Computer Science and Engineering

CGPA: 9.3 / 10.0

Dec 2020 – May 2024

Coursework

Artificial Intelligence and Expert Systems Soft Computing Techniques Pattern Recognition Big Data

Certifications

AI for Everyone

DeepLearning.AI

Supervised Machine Learning

Regression & Classification

Advanced Learning Algorithms

DeepLearning.AI

Unsupervised Learning

Recommenders & RL

Machine Learning Specialization

Standford / DeepLearning.AI

Build with AI

Create an Agent with GPT-OSS

Professional Experience

Microchip Technology Inc.

Intern | May 2026 – Aug 2026

Roseville, CA, USA

  • Engineered a cross-platform Python GUI to interface with custom SSD firmware for Future of Memory and Storage (FMS) 2026, driving technical demonstrations of two storage technologies to 500+ clients.
  • Automated PCIe/NVMe hardware benchmarking for AI workloads by writing Python scripts to stress-test evaluation boards across 1000+ iterations, identifying firmware bottlenecks and debugging complex customer edge cases.

HCL Tech

Technical Lead | Sep 2024 – Jul 2025

Madurai, Tamil Nadu, India

  • Trained in Microsoft Azure services, including App Service, Azure SQL Database, and Azure DevOps, gaining expertise in cloud migration and deployment pipelines.
  • Contributed to migrating a legacy corporate website (50K+ monthly users) from on-premise servers to the Azure platform, improving availability and reducing infrastructure costs by an estimated 20%.

National Institute of Ocean Technology (NIOT)

ML Research Intern | May 2023 – Jun 2023

Chennai, India

  • Performed extensive exploratory data analysis (EDA) on 1M+ meteorological, oceanographic, and ship trajectory data points to identify key environmental factors impacting fuel efficiency and optimal navigation.
  • Developed a robust data preprocessing pipeline to clean, integrate, and normalize large raw datasets (CSV-to-Excel transformation, missing value handling, feature scaling) to enable reliable downstream analysis.

Projects

VibeArchitect

Gemini API • FLUX.1 • Full-Stack Agentic AI

Agentic Content Pipeline

Special Mention, TrojanHacks USC

Automates multimodal content trend analysis and thumbnail generation using Gemini and FLUX.1 with 100% system uptime.

Gemini APIHuggingFaceReactNode.js
View Project

Vision Routing Engine

PyTorch • YOLOv11 • Object Detection

Dynamic Vision Pipeline

Cascades X-ray screening between NanoDet-Plus and YOLOv11L, slashing inference time by 54% while improving AP@50 by 6.9%.

PyTorchYOLOv11NanoDet-PlusCV
View Project

Kuri AI

Multimodal LLMs • Agentic AI • Supabase

LLM Task Understanding

Delivers multimodal intent classification and hierarchical task scheduling with Google Calendar and Supabase serverless functions.

Gemini APIMultimodal LLMsSupabaseVercel
View Code

Valorant RAG Chatbot

Semantic Search • ChromaDB • Mistral 7B

Valorant Strategy Assistant

Delivers strategy-focused gaming Q&A using ChromaDB semantic retrieval and a quantized 7B Mistral LLM running 30% faster.

Mistral 7BRAGChromaDBPython
View Code

Pix2Pix Maps

GAN • Computer Vision

Aerial Image to Map

Translates complex aerial satellite imagery into high-fidelity cartographic map representations using conditional GANs.

PyTorchTensorFlowGANsCV
View Code

X-Ray Vision

CNN • Medical Imaging

COVID-19 Classification

Classifies chest X-rays via a two-stage deep CNN pipeline, attaining 91% abnormality accuracy and 96% disease specificity.

CNNKerasPython
View Code
🏆 Quidditch Seeker Mode — Catch the Golden Snitch!