Ashwini Deshmukh

About Ashwini Deshmukh

I enjoy building AI systems that actually solve real problems, not just demos that look good in presentations.Over the last 11+ years, I’ve worked across data science, data engineering, and analytics, and more recently focused deeply on Generative AI, LLMs, and RAG-based systems. I like working on problems where scale, accuracy, and real-world usability all matter.In my recent work:• Built an LLM-based chat summarization system using RAG that reduced manual investigation effort by 60%• Worked on improving LLM reliability using evaluation frameworks like DeepEval (especially around hallucination and relevance)• Explored Agentic AI approaches for next-best-action and autonomous decisioning• Automated reporting workflows using Azure dashboards and Power Automate, cutting down manual effort by ~70%Earlier in my career:• Led GenAI use cases at PwC using Azure OpenAI• Built data pipelines using Airflow and BigQuery for large-scale analytics• Worked on fraud detection and credit risk models with measurable business impactI’m particularly interested in:Generative AI, LLM evaluation, RAG systems, Agentic workflows, and scalable data platforms.At the core, I enjoy translating business problems into practical AI solutions that people can actually use.Open to Senior Data Scientist / GenAI Engineer / AI Lead opportunities.

Experience

1. Gen Ai Specialist

   IBM

   Sep 2024 — Present

Education

• Vignan Bharathi Institue of Technolgy

  Bachelor of Technology - BTech, Computer Software Engineering

  2008 — 2012