Services · SAP BAIP

Extend the core. Do not modify it.

The Business AI Platform is where the interesting work now lives — and where the next generation of technical debt is quietly being written. We build extensions that survive the upgrade.

The white space

Clean core is a discipline, not a license.

Moving custom code off the core does not make it governed. It makes it someone else’s problem, on a platform with fewer guardrails.

Where it goes wrong

Three ways the platform fills up

01

Extension without an owner

Six teams building on BAIP with no shared data contract, and no one able to say which service is authoritative.

02

Integration by accumulation

Point-to-point flows added one project at a time, until nobody can draw the picture on a whiteboard.

03

AI on ungoverned data

A model trained on the copy of the copy, answering confidently and citing nothing.

What we do

Five practices, one architecture

01

Strategy and roadmap

What belongs in the core, what belongs on BAIP, and what should not be built at all. Written down, with the reasoning.

02

Application development and extension

Side-by-side extensions using CAP and Fiori, built to the extensibility model rather than around it.

03

Data management and analytics

Governed replication, a named source of truth per object, and analytics that can show their working.

04

Integration

Integration Suite used as an architecture, not as a connector shop. Event-driven where the business is event-driven.

05

Intelligent technologies

AI, machine learning and IoT applied where the underlying record is already defensible — and honestly refused where it is not.

How we work

Governance first, code second

01

Map what is already on the platform

Including the parts that have not come up in a steering meeting yet.

02

Agree the data contracts

Which system owns which object, and what downstream is allowed to assume.

03

Set the extension pattern

One pattern, documented, that every team builds to.

04

Build and hand over

With the tests, the runbook and the upgrade path, not just the code.

Next step

Start with the hardest part

A short conversation about where your SAP, GIS and schedule data stop agreeing — and what it would take to bring them into line.