Implementing AI agent for better procurement decision-making for a major pharma company

About our customer

The scenario – Challenges with manual processes

The indirect procurement team handled 300–400 purchase requisitions (PR) per day across multiple categories. While SAP was the system of record, buyers relied heavily on manual data gathering and email-based coordination to process each request.

This created multiple pain points:

Buyers spent significant time collecting information, like supplier history, pricing benchmarks, quality data, and past orders, before acting on each PR.

High PR volumes led to buyer bottlenecks, extending PR-to-PO cycle times.

Approvals routed over email increased delays and made DOA compliance difficult to enforce consistently.

Decision-making was reactive, with limited ability to prioritize high-impact PRs or identify savings opportunities early.

Want to share this story with your peers?

Our solution – an AI agent to streamline PR-to-PO process

Saxon implemented an AI procurement agent that accelerated the PR-to-PO cycle by removing manual buyer effort and embedding decision support directly into procurement workflows.

The agent automatically processes PRs and helps buyers prioritize requests and make informed supplier decisions. The agent speeds up the PR-to-PO process by automating the multi-level approval workflow, while maintaining strict adherence to delegation of authority (DOA) framework.

Buyers can interact with the agent in their natural language and execute the process from Microsoft Teams itself, without switching between multiple apps.

Deep integration

The agent is hosted on Azure and integrated with the customer’s SAP, Microsoft Teams, Outlook, and SharePoint systems for seamless data flow across systems