ABOUT

Procurement, Made Conversational

A global energy organisation had adopted a new procurement platform, but requestors found the experience complex, click-heavy and inconsistent from step to step. With the platform already live and owned by an external team, any improvement had to be realistically buildable and shippable.

​

Partnering with frog, I helped reimagine the request-to-order journey as an AI-assisted experience. Document-to-request automation extracts key information from an uploaded file and pre-fills the request. Chat-led, agentic flows let users describe what they need while the system assembles the request, and people stay in control through a final review step that works as an index, letting them jump back to any phase.

​

Stakeholder research revealed two user profiles: experts who move fast, and occasional users who need guidance and confirmation. This became the organising principle for AI personalisation. Speed channels serve experts, while added guidance and confirmation support newcomers, across the chat flow, navigation, progress and line items.

​

To keep the vision deliverable, every ideal concept was paired with a simpler technical fallback. Concepts were validated in close collaboration with the platform team and aligned through regular sessions with users.

​

The result was a prioritised, feasibility-checked set of design concepts, handed over ahead of launch and tied to the client’s development roadmap.

​

This project is under NDA and further details cannot be shared online.