Case Studies

Advanced GenAl Solutions Reduced Extraction Time

Written by KPI Partners News Team | Apr 30, 2024 2:30:46 PM

The Objectives

The client sought an advanced GenAI-driven solution to efficiently manage complex rebate contracts, aiming to improve data extraction, interpretation, and transformation for enhanced business operations.

 

Challenge

  • Complex Contract Diversity: Rebate contracts varied widely in format, complicating management across different buying groups and vendors
  • Non-standardized Terms: Contracts used a variety of keywords and terms to express similar values, lacking standardization.
  • Tiered Data Extraction: The tiered structure of rebate contracts made data extraction from multiple sections challenging..
  • Manual Data Conversion: Converting complex contract terms into Excel files manually was error-prone and inefficient.
  • Data Interpretation Difficulty: The complexity of contracts requires a sophisticated approach to accurately interpret and map data.

 

Solution 

The solution architecture involved a multi-faceted approach using AWS services to transform the client's rebate contract management system.
  • AWS OCR Technology: Utilized for the initial extraction of key-value pairs, numerical values, and terms from the contracts.
  • AWS Bedrock GenAI: is Customized and trained to understand the contract structure, enabling intelligent comprehension and processing.
  • Data Conversion: Transformed extracted terms and conditions into data values that matched manual entries, ensuring consistency.
  • Data Mapping: Aligned the converted data with the Kariba data model structures, facilitating integration into the client's systems.
  • Vendor Rebates Dashboard: Integrated the data model to support the dashboard, providing a user-friendly interface for data analysis.
  • AWS Textract: Employed to convert contract text into structured datasets, enhancing data organization and accessibility.
  • AWS OpenSearch: Indexed the extracted data, enabling fast and efficient search and retrieval capabilities.
  • AWS Bedrock: Interpreted organizational terminology and transformed data into the reporting data model with the help of a metadata layer.
  • Solution Deployment: The entire solution was integrated and deployed to effectively support the client's operational needs. Continuous Improvement: The solution was designed with scalability, allowing for future enhancements and updates as needed.

Solution Architechture

 

 

Impact

  • Extraction Time Reduced: The use of advanced AI solutions led to a 90% reduction in the time required to extract data from complex contracts
  • Error Rate Decreased: The accuracy of contract data extraction improved, with a 30% reduction in errors.
  • Scalability for Growth: The AI-driven solution was designed to handle large volumes of contracts, supporting the organization's growth.
  • Historical Record Keeping: Enhanced tracking and reference capabilities facilitated auditing and compliance.
  • Operational Efficiency: The overall efficiency of managing rebate contracts was significantly improved, benefiting the organization's operations.