Piyush Sonawane

Data Analyst @FNF India Private Limited

Bengaluru, IN

Project: Automated Document Extraction for U.S. County RecordsI’m currently working on an end-to-end document extraction POC for U.S. county-level public records. These documents come in multiple formats such as Deeds of Trust, Mortgages, Liens, and Starters, each with its own structure and terminology.Our objective is to automate the extraction of all business-critical fields from these documents using Large Language Models (LLMs) available through AWS Bedrock and Azure AI Foundry.This involves designing robust prompt workflows, Researching different prompt formatting like YAML, Markdown, POML by Microsoft etc, and ensuring high accuracy across diverse document types.