SAP Data Quality · Migration · Archiving · Analytics

Your SAP system already knows the answer.

Twenty-six years inside enterprise data. I make SAP data trustworthy enough to make decisions on — and clean enough for AI to actually use.

Singapore-based · Serving Singapore, APAC and European companies with regional operations

26Years in enterprise data analytics
30+SAP customer projects delivered
2018Stanford-certified in machine learning
2026SAP-certified Business Data Cloud (SAP BDC)
2027SAP ECC mainstream maintenance ends
Why this matters now

Most SAP data problems only become visible when it is expensive to fix them.

Three things are converging on the same eighteen months. Each one is survivable alone. Together they are the reason migration budgets double and AI pilots quietly disappear.

01

Your AI pilot stalled — and it wasn't the model

Enterprise AI rarely fails on algorithms. It fails on duplicated master data, undocumented lineage, inconsistent units, and nobody being able to say where a number came from. The fix is upstream of the model, in data most organisations have never profiled.

02

2027 is closer than your migration plan assumes

SAP mainstream maintenance for ECC ends in 2027. The technical migration is the easy part. The hard part is twenty years of accumulated data — what to carry, what to clean, what to archive, and what should never have been there at all.

03

Nobody knows what is inside the Z-tables

Custom objects, bespoke ABAP, workarounds built by people who left years ago. Every SAP estate carries this. It is invisible until a migration or an AI project reaches for the data and finds it cannot be trusted.

How the work pays for itself twice

The migration and the AI programme are the same programme.

Most organisations run these as two initiatives with two sponsors and two budgets, and do the foundational work once, badly. It is one body of work with two payoffs.

Where you are Your SAP estate 20 years of accumulated data, custom code and undocumented Z-tables Step one Profile and measure Quality baseline, duplicate counts, code usage, lineage — evidence instead of assumption Step two Remediate and reduce Fix what matters, delete what is unused, archive what must be kept but not carried Payoff one A migration that holds S/4HANA scoped on evidence, with the surprises found early and cheaply Payoff two AI you can defend Known lineage, agreed definitions, governed data — the prerequisites, in place

Swipe to see the full diagram →

Your data quality baseline sets your AI ceiling. The migration is the cheapest opportunity you will get to raise it, because someone else is already paying for the disruption.

Engagements

Fixed scope. Fixed price. A written answer you can circulate.

Every engagement produces a document your team can act on without me in the room. No open-ended discovery, no day-rate drift.

01
10 working days

SAP Data Readiness Check

Before you commit seven figures to S/4HANA, know exactly what condition your data is actually in — not what the documentation claims.

  • Data quality scorecard across your critical master and transactional objects
  • Custom object and Z-table inventory, with usage analysis
  • Master data duplication and consistency assessment
  • Archiving and decommissioning candidates, quantified
  • Risk register and a costed remediation roadmap
02
15 working days

AI Foundation Audit

Find out whether your data can support the AI use cases your board is asking for — before you spend the budget finding out the hard way.

  • Feasibility assessment for each proposed use case, against your actual data
  • Data lineage map: where each number originates and what happens to it
  • Governance and PDPA gap analysis
  • A prioritised shortlist of use cases that are genuinely buildable today
  • One working proof-of-concept on your own data
03
4–6 weeks

Hidden Margin Finder

Your operational systems do not talk to each other, and money leaks in the gaps between them. I connect them and show you where — with numbers attached.

  • Unified data model across your operational and financial systems
  • Leak analysis with quantified impact, ranked by size
  • A live dashboard your team keeps after the engagement ends
  • Particularly suited to hospitality, F&B and multi-site operations
04
Rolling · 3-month minimum

Fractional Data Architect

Architect-level judgement four to six days a month, without carrying the headcount. For teams who need the seniority but cannot justify the salary.

  • Architecture and design review on live programmes
  • Vendor and system integrator oversight — a technical second opinion
  • Mentoring for your internal data team

Singapore companies: most of this work is co-fundable

Government schemes co-fund capability and technology projects for eligible SMEs. We scope engagements with that in mind and provide the proposal and quotation the application requires.

Read the funding guide
Who you would be working with

I have spent twenty-six years inside other people's data.

I am Bernd Murr. I have worked in enterprise data analytics since the late 1990s, across dozens of SAP customer projects, and I moved to Singapore to do the same work here. Murr Analytics & AI is a Singapore private limited company.

My background is unusual in one specific way: I am equally comfortable in the deep plumbing of an SAP system — ABAP, table structures, twenty years of accumulated custom development — and in the machine learning layer that increasingly sits on top of it. I studied how AI systems actually work in 2018, with certification from Stanford University, well before it became a procurement category.

That combination matters more than it used to. The organisations getting real value from AI are not the ones with the best models. They are the ones whose data was already in order.

You work with me directly. No account manager, no rotating junior team, no handover to someone who has not read the brief.

Bernd Murr, founder of Murr Analytics & AI
Bernd Murr · Founder Singapore

Technical depth

SAP

SAP Business Data CloudSAP Datasphere S/4HANASAP ECC ABAPSAP BW HANA Data MigrationArchiving Master Data

Data & AI

PythonSQL Machine LearningData Lineage TensorFlowPyTorch Data GovernanceAnalytics

Also

JavaJavaScriptPDPA

Working with European headquarters

I work in English and German. For European groups with APAC operations, I can bridge the gap between headquarters reporting standards and what the regional entity is actually able to produce — a problem that is usually described as a systems issue and is almost always a data one.

Free · No sales call required

The 2027 SAP Data Readiness Report

A practical guide for Singapore and APAC companies on what to do with their SAP data before mainstream maintenance for ECC ends. Written for people who will have to do the work, not for people approving the budget.

  • What actually breaks during an S/4HANA data migration
  • How to inventory twenty years of custom objects
  • Deciding what to migrate, what to archive, and what to leave behind
  • Why your data quality baseline determines your AI ceiling
  • A pre-migration data checklist you can run yourself
Cover of The 2027 SAP Data Readiness Report

24 pages · 12 sections · no registration wall beyond your email

Get in touch

Start with a conversation, not a proposal.

Thirty minutes, no obligation, no slides. Tell me what you are trying to do and what is in the way. If I am not the right person, I will usually know someone who is.

  • Email
  • LinkedInBernd Murr
  • Based inSingapore
  • LanguagesEnglish · German
  • ServingSingapore & APAC · European groups with regional operations