Projects.

What I have been working on in recent years.

A selection. More projects to follow.

Data integration

Flexible data loading as SaaS on SAP BTP

Combine any source with any target, the app takes care of the data load.

editor / extending-rap-apps-cleanlyPublish

Extending RAP apps cleanly

Why extensions break

The way through the base layer

An example from the project

What stays of it

Extending RAP apps cleanly

Blog · draft 3 · last edited 8 minutes ago

An extension that starts at the wrong place lasts exactly until the next release. That is rarely down to the code and almost always to the layer it touches.

Why extensions break

Whoever overrides the behaviour directly inherits every change of the standard. The detour through the base layer costs a day and holds for years.

Shorten the intro.

Cut to three sentences. Cover image done, graphic running.

Cover imageGraphic

Blogs, images, LinkedIn posts and more with AI

Generating content with AI: from creation to publication, all in one app.

Z_REVENUE · Translation

Z_REVENUE.abapZ_FX_CONVERSION.abapZ_DOC_STATUS.abap

LOOP AT result_package
     ASSIGNING <fs>.
  SELECT SINGLE rate
    INTO lv_rate FROM zrate
    WHERE waers = <fs>-waers.
  <fs>-amount = <fs>-amount
                * lv_rate.
ENDLOOP.

AI · TranslationApply

Z_REVENUE to Datasphere

The routine fetches the rate per record at the key date. In the view this becomes a join on the rate table with the key date as a parameter.

Different from BW

Rate via join instead of SELECT SINGLE per row

Key date as parameter :p_keydate instead of lv_keydate

DeviationRounding: ABAP rounds to 2 places, the view keeps the raw value

SQL · Datasphere

SELECT t.belnr, t.waers,
  CASE WHEN t.waers <> 'EUR'
    THEN t.amount * k.rate
    ELSE t.amount END AS amount
  FROM revenue AS t
  LEFT JOIN zrate AS k
    ON k.waers = t.waers
   AND k.datum <= :p_keydate

Line 5, column 24SQLUTF-81 note open

Migrate BW logic to Datasphere with AI

The app reads legacy routines, translates them, a review marks every deviation, as SaaS on SAP BTP.

Data validation with AI

Rules from the data, rules by chat, checks across systems: as SaaS on the SAP BTP.

SAC chatbot with access to story, model and SAP system

Analysing, importing, writing back and starting retractions, in natural language and without switching tools.

SAPBed blocksSF

WardAllPeriodSept 12 to Sept 30StatusblockedGoAdapt filters

Blocks (18)Record

BedWardReasonFromToStatus

073BRepair11.09. 14:0015.09. 12:00blocked

122AIsolation12.09. 06:0016.09. 10:00blocked

044CCleaning13.09. 09:0014.09. 12:00blocked

091BStaff10.09. 07:0020.09. 07:00blocked

024CRepair09.09. 13:0014.09. 16:00blocked

112ACleaning08.09. 08:0013.09. 18:00free

Ward 3B · 12 beds · 2 blocked

From12.09.2026, 08:00

To14.09.2026, 14:00

Business application

Blocked beds recorded in one place

Which bed is blocked and why is in one app instead of in lists. Reporting included.

SAPEvaluationsSF

Evaluations (24)

BEW-0341Annual evaluation · Asset 4711in progress

BEW-0340Interim evaluation · Asset 4708released

BEW-0339Annual evaluation · Asset 4702released

BEW-0338Special evaluation · Asset 4699in review

BEW-0337Annual evaluation · Asset 4690released

Annual evaluation · Asset 4711

BEW-0341 · Technology · in progress

Area:Technology

Reference date:30.09.2026

Reviewer:Department head

Value must be between 0 and 100

Process digitisation

Evaluation workflows in one SAPUI5 app

Entries are checked in the frontend and in the backend and go straight into the system. Without an Excel list.

Process digitisation

Repair orders recorded digitally

200 orders a day, checked as they are entered and automatically in S/4. Without an Excel list.