Projects

AI systems from idea to production.

Each project highlights a full pipeline: data readiness, model development, deployment, and measurable business impact.

Portfolio

Selected deliveries

Recommender System

Recommender System

Hybrid model with SVD, KNN, neural nets, and GPT-4 explainability. Achieved 70%+ hit rate across 5 industry sectors.

Personalization RAG MLOps
Churn Prediction

Employee Churn Prediction

Interpretable XGBoost model with 50+ features and LLM sentiment signals, predicting churn 6 months ahead for proactive retention planning.

XGBoost Explainability
Anomaly Detection

Anomaly Detection

Ensemble framework with PCA and logistic regression risk scoring, expanding monitoring coverage by 35% across small-revenue projects.

Risk Monitoring
Time Series Forecasting

Demand and Supply Forecasting

AutoTS with reconciliation; achieved 20% higher accuracy and automated monthly workforce planning.

Forecasting SHAP

RAG Data Discovery Assistant

GPT-4o powered RAG system indexing 100+ data sources, reducing exploration time by 20%.

Azure AI Search LangChain

AI Agent Framework

Multi-agent system for email drafting, summarization, and meeting notes with LangGraph and PromptFlow to improve organizational productivity.

LangGraph Agentic AI