Hi, I'm Harish Nandhan

Real-time supply chain monitoring and ML-powered delivery prediction system. End-to-end data pipeline extracting shipment data from Google Sheets, processing through Kafka and MongoDB, transforming with dbt, storing in BigQuery.

💡 Complete streaming pipeline with ML-powered delivery predictions

Real-time sentiment analysis streaming pipeline that processes and analyzes text data for sentiment classification at scale.

💡 Live sentiment monitoring and analysis

Lightweight real-time fraud detection and alerting prototype using Kafka, MongoDB, and Python. Streaming pipeline that produces synthetic transactions, evaluates them with fraud rules, and sends email alerts.

💡 Real-time fraud detection with automated alerting

Comprehensive collection of my learnings and implementations in Data Science and Analytics

💡 Structured knowledge base for DS concepts

Task Insights and Workload Overview Analysis for Denver International Airport

💡 Optimized task allocation and resource planning

Sales & Finance Analytics Reports for Atliq Hardware

💡 Data-driven insights for business decisions

Intelligent voice-powered assistant for querying SQL databases using natural language. Speak your question, and the app transcribes speech, generates SQL using LLM, executes it, and displays results with automatic visualizations.

💡 Natural language to SQL with voice interface

Intelligent medical diagnostics assistant powered by EuriAI and LangGraph that analyzes symptoms and provides preliminary diagnostic suggestions with treatment recommendations.

💡 AI-powered preliminary medical diagnosis

Powerful RAG chatbot for uploading PDF documents and asking questions about their content. Built with Streamlit, FAISS for vector search, and EURI API for embeddings and chat completions.

💡 Intelligent document Q&A with RAG

Performance comparison and benchmarking analysis of various Large Language Models

💡 Systematic evaluation framework for LLM selection

AI-powered medical chatbot using LLAMA 2 for healthcare queries

💡 Accessible medical information assistant

Transformer-based model for generating educational videos from text

💡 Automated educational content creation

Deep learning model for medical image segmentation to detect pneumothorax

💡 Automated detection for faster diagnosis

Predicting Body Mass Index using facial image analysis with deep learning

💡 Non-invasive health metric estimation

Deep learning healthcare project for pneumonia detection from chest X-rays

💡 Improved diagnostic accuracy

ML model to predict student academic performance based on various factors

💡 Early intervention for at-risk students