• Developed custom pipelines to ingest and preprocess structured and unstructured datasets from multiple state sources.
• Integrated large language models (LLMs) to generate interpretive summaries and highlight anomalies in economic, healthcare, and education statistics.
• Optimized real-time data querying, reducing manual analysis time by over 60%.
• Ensured model explainability and compliance with state data governance policies.

Impact:
Enabled fast, accurate, and AI-assisted decision-making for state-level planning and policy evaluation.