Enterprise & SAP · Portland AI Works

What I do

Three kinds of engagement.

SAP + AI integration

AI steps inside existing SAP processes, built the way the rest of your landscape is built. SAP Integration Suite, S/4HANA, Clean Core extensions, Anaplan. Approval flows, document processing, exception handling, master data hygiene.

MCP-first system design

Systems where an AI agent and a person operate through the same commands, the same validation, and the same audit log. saybooks.io is the reference implementation. I design and build this pattern for your systems.

Fractional AI architect

A senior engineer on a fixed number of days a month. Reviews designs, unblocks the team, keeps vendor claims honest, and builds the parts nobody else has time to build.

How I think about it

Your AI should work inside your system of record, not beside it. A model that writes to a spreadsheet is a demo. A model that posts a validated document through the same interface a clerk uses is a system.
Clean Core is not a constraint on AI, it is the reason AI is safe to add. Side-by-side extensions with proper APIs are exactly where agents belong.
Every agent action is logged with the same actor, reason, and timestamp fields as a human action. Auditors should not need a separate story for the AI.
Fixed scope when the scope is known. Day rate when it isn't. Never an open-ended program with no end date.

Pricing

Day rate or fixed scope, quoted after a 30-minute call. Remote by default, on site when the work calls for it.

Book a 30-minute call LinkedIn