Hemanth Bantwal

Technology leader in commercial lending platforms and applied AI, and a PhD researcher on graph-based retrieval for regulated document workflows. Based in Pittsburgh, Pennsylvania.

My work sits where large language models meet documents that carry legal and financial weight: loan agreements, credit memos, payer policies and prior-authorization forms. The common thread is building retrieval systems that can show their work, so a model's answer can be traced back to the clause or policy it came from.

I lead commercial-lending technology at a large US bank, build AI solutions for financial-services clients at a global consultancy, and am completing a PhD on GraphRAG evaluation. The GraphRAG Architecture Gallery is my running field guide to the systems in this space.

Now

• Benchmarking graph-based versus vector retrieval on commercial-lending documents (open dataset in preparation).
• A knowledge graph of payer policy for prior-authorization documentation.
• Speaking and reviewing in the knowledge-graph and AI-in-finance communities; see Talks & recognition.

Background

• 2023 – present
  Associate Director of Technology, Synechron — AI solutions for financial-services clients
• 2017 – present
  Commercial Lending technology leadership, PNC — origination, servicing and booking platforms
• 2012 – 2023
  SDLC Partners — consulting on lending and healthcare-payer platforms
• 2012 – 2016
  Highmark / HM Health Solutions — payer integration and operational readiness
• 2008 – 2011
  Infosys — systems engineering
• PhD (in progress)
  Robert Morris University — GraphRAG and knowledge graphs for regulated documents

Edit this page in about.html. Keep it to what you would say on a conference bio; the private evidence vault holds the detail.