Sohini Das

I’m an Applied AI Engineer with 3+ years delivering scalable AI/ML, cloud data engineering, and Generative AI solutions. I focus on turning data into reliable models and real-world workflows—especially where enterprise knowledge and automation matter.

I’ve built and deployed Retrieval-Augmented Generation (RAG) applications using Azure ML, combining LLMs, prompt engineering, and context-aware retrieval to support knowledge-grounded AI document generation. I also developed demand forecasting models using feature engineering, ensemble techniques, and rigorous evaluation, reaching ~86.6% prediction accuracy.

On the data engineering side, I’ve worked across Azure Data Factory, Azure Databricks, and SQL Server, and I’ve engineered reusable pipelines that improve quality and accelerate experimentation. Notably, I reduced Azure ETL costs by 25% through ETL/pipeline optimization and improved operational reporting workflows by ~65% via automated Python–MySQL reporting.

I enjoy building end-to-end solutions—from preprocessing and model development to deployment and CI/CD-ready engineering practices. I bring a practical, delivery-focused mindset across analytics, machine learning, and Generative AI, partnering with teams to ship dependable intelligence and automation.