A new data foundation, with Claude built in from day one

The Data Forest

September 7, 2026

About ASAP HR Group

ASAP HR Group is a Belgian staffing and HR services company based in Genk, providing payroll and workforce services to hundreds of external clients. ASAP is migrating its payroll system from Prato HiAnt to Earnie, with a strict go-live deadline.

The challenge

Calculation logic sat in undocumented database scripts, so no one could verify with confidence how a number was actually produced.
The reporting platform wasn’t built to support a payroll system switch, or to scale client reporting from a few dozen customers toward 300–400.
Rolling out Claude on top of an ungoverned data foundation would only produce fast, confident, wrong answers.
ASAP needed a way to bring AI into the organization without creating new risks around sensitive HR and payroll data.
ASAP needed a working platform and a working AI rollout, both on a fixed deadline.

Our approach

1. Strategy and platform decisions

We mapped ASAP’s data landscape and agreed the target architecture, tooling stack and delivery approach. Part of that decision was putting Claude into the development workflow from the first commit, a call that also helped make the internal case for rolling out Claude licences company-wide.

2. Platform build, with Claude in the build itself

We built the ingestion, transformation and reporting layers: automated daily loading of Earnie, the legacy system and other sources into one place, calculation logic moved out of hidden scripts into version-controlled and tested code, and Power BI reporting on top. Claude was deployed to ASAP’s IT staff and used directly in the build, not introduced afterward through training.

3. Review pipeline and adoption rules

Claude now reviews and tests every data model change in the CI/CD pipeline before it reaches production. Alongside the build, we trained ASAP’s team to run and extend the platform on their own.

Specific challenges we solved

Payroll logic hidden in scripts: Moved into version-controlled, tested code, with every change reviewed before it goes live.
No shared definitions: Internal and client reporting now run on the same agreed data definitions and governance rules.
No review process for changes: Claude tests and reviews every pull request before it reaches production.

“The Data Forest delivered exactly what they said they would: a modern AI-integrated data platform. The Data Forest is the kind of partner that helps you scale for the digital future.”

— CFO, ASAP HR Group

Results and benefits

The new platform foundation is live in production, replacing undocumented logic with version-controlled, tested code.
The new platform was built 67% faster by using AI (Claude) versus the old traditional way (4 months vs 12 months).
Every data model change is reviewed and tested automatically before release.
Claude is adopted company-wide and used directly in engineering work, backed by a clear usage policy.
ASAP’s engineers can ship changes to production without The Data Forest involved.
The Data Forest and ASAP are moving into the next phase: migrating payroll data onto the new platform ahead of the January 2027 switch.