CK Bhatia

Attendees will see how a lightweight, chat-driven interface can guide staff through loan processing while automatically validating and organizing data from scanned documents. This practical, agentic approach shows that real AI transformation is possible even with

limited resources.

• Laying the Foundation: Identifying impactful, low-cost AI opportunities to automate internal workflows.
• Using OCR and LLMs to digitize and extract critical loan data from unstructured PDFs.
• Agent-Driven Workflows: Enabling staff to manage and validate loan data through an intuitive chat-based interface.
• Building on a Budget: Leveraging off-the-shelf models and no-code/low-code orchestration to stay cost-effective.
• End-to-end showcase of a conversational system that handles document ingestion and variable extraction in real time.
• Maintaining data privacy, compliance, and trust without a heavyweight AI risk team.