© Copyright zerodays d.o.o.
For a large Slovenian pharmaceutical company we developed an AI-driven internal compliance solution that streamlines the process of evaluation of supplier proposals against Requests for Proposals (RFPs). The system automates the extraction, comparison, and scoring of supplier data.
Suppliers typically respond with complex documents exceeding 60 pages, containing both structured and unstructured data. Our AI agent analyzes these offers in depth and accurately maps all relevant information to the corresponding RFP requirements. The agent also links each result to the exact quote in the document, so users can easily cross-check and verify it themselves.
Every line item from each supplier’s offer is checked individually. The agent determines whether a requirement is compliant, non-compliant, or not found, and assigns a confidence score indicating how clearly the requirement is defined and matched.
The system automatically generates a clear, side-by-side overview of the best proposals. What previously required manual comparison of 5+ offers now takes less than 15 minutes, down from 9 hours.
An upgrade to the first version, V2 introduces a fully configurable data layer: any table structure can be added to the system, along with customized processes tailored to each team's way of working. Every attribute can be traced back to its origin, with comprehensive access logs (audit logs) and version history — enabling a tabular, spreadsheet-like view — so users always know where the data came from and how it has changed. The file and project structure is also fully configurable, allowing teams to adapt the system to their own way of working, instead of having to adapt to a predefined template.
