I am a Data Scientist and AI Engineer passionate about building intelligent systems that solve real-world problems. With a Master's in Data Science from Indiana University and a strong engineering background, I specialize in Generative AI, Agentic Systems, LLMs, and Full-stack Development.
π Location: Washington D.C.
π Education: M.S. in Data Science (Indiana University Bloomington)
πΌ Open to work: Actively seeking full-time opportunities in AI/ML and Data Science.
- Languages:
PythonJavaScriptC++SQLR - AI/ML:
PyTorchLangChainRAGAgentic AICrewAIGemini/OpenAI APIsComputer VisionNLP - Web & Backend:
FastAPIStreamlitFlaskReactNode.js - Cloud & DevOps:
AWSDockerKubernetesCI/CDMLflow - Data:
PostgreSQLMongoDBPineconeFAISSKafkaSpark
| Project | Description | Tech Stack |
|---|---|---|
| Multi-Agent AI Travel Planner | An autonomous multi-agent system that orchestrates specialized agents (Flight/Hotel Analysts) to generate personalized travel itineraries. | Python CrewAI Gemini 2.5 FastAPI Streamlit SerpAPI |
| Agentic AI for Supply Chain | An intelligent RAG system for international supply chain management, enabling natural language querying of contracts and shipments. | LangChain LangGraph MongoDB Atlas Claude 3.5 Tavily API |
| Reddit Thread Analyzer | A multi-modal GenAI application that analyzes and summarizes lengthy Reddit threads, including image content verification. | Python Gemini Vision AsyncIO AWS S3 Streamlit |
| AI-Powered Frisbee Rules Assistant | Full-stack RAG assistant for querying complex frisbee rules with source attribution. | LlamaIndex FastAPI React PostgreSQL Ollama Docker |
| Contextual Document Chatbot | A RAG-based QA system capable of real-time querying over custom document corpora with high retrieval accuracy. | LangChain OpenAI GPT-3.5 Pinecone FAISS |
| Project | Description | Tech Stack |
|---|---|---|
| Unsupervised Learning for Text | Analyzed 52k+ Reddit comments using clustering and topic modeling to quantify discourse themes and sentiment. | Scikit-learn NLP K-Means LDA VADER Matplotlib |
| Traffic Sign Classification | Real-time traffic sign recognition system achieving 98% F-1 score, deployed as a web app. | PyTorch MobileNetV2 Computer Vision Streamlit |
- π§ Email: nainesh.h.rathod@gmail.com
- π LinkedIn: linkedin.com/in/nainesh-rathod
- π Portfolio: github.com/nainesh-20