Projects
Live from GitHub — all my public repositories across applied machine learning, generative AI, data science, and full-stack development. Updates automatically whenever I push new work.
9 public repositories · auto-refreshed every 24 hours
AdvocateAI
RAG-powered legal matching platform — semantic search with vector embeddings (FAISS/Qdrant), LLM-assisted recommendations, FastAPI backend, and Flutter frontend. Deployed on DigitalOcean.
Smart CSV Toolkit
Smart CSV Toolkit — LLM-Assisted CSV Cleaning & Metadata Inference. A Streamlit app to clean, merge, and analyze CSVs with guided steps, metadata inference, interactive decision trees, and optional LLM support.
Autoencoder Optimization Anomaly Detection
Convolutional autoencoder-based anomaly detection vs shallow baselines — IEEE IJCNN 2024. Live demo on HuggingFace Spaces.
EDA Hypothesis Regression
Three end-to-end R projects: EDA on demographics, hypothesis testing on smoking & birth weight, and linear regression on Seoul bike-sharing demand.
Ab Lab
🧪 AB Lab is an end-to-end A/B testing toolkit built in Python with an interactive Streamlit interface. It enables researchers, data scientists, and product teams to design, simulate, and analyze experiments—from power calculations and bias reduction to frequentist and Bayesian inference.
Geospatial
Production geospatial ML platform - satellite classification, anomaly detection, change detection, tree crown segmentation, and LiDAR forest inventory. FastAPI + PyTorch + Railway.
Insightdesk Ai
AI-powered IT helpdesk platform — ticket categorization, RAG-based solution retrieval, anomaly detection & model monitoring. Built with FastAPI + React + Tailwind.
AirKube
Agentic MLOps platform — LangGraph + Gemini AI agent, Neo4j Knowledge Graph, FastAPI inference API, Airflow pipelines, and Datadog observability.
Geniebot
GenAI Telegram Bot with RAG-based Q&A and Image Captioning using local LLMs (Ollama) and persistent vector DB