- 10+ years
- of data migration & MDM
- Atlas Copco · Epiroc · Åhléns
- Global ERP & MDM programmes
- AI · MDM · ERP
- AI agents, master data & ERP platforms
Three areas – one foundation in governed data
AI agents are only as good as the data they work with. That's why my three areas belong together: master data gives one shared truth, migration moves it correctly – and AI makes the work faster, with people still in control.
AI Engineering
AI agents and LLM solutions that do real work in the business – under human control and fully traceable.
- Custom MCP servers connecting language models to ERP, MDM and business data
- Multi-agent workflows with clear roles, budgets and human approval
- Commercial models via API or self-hosted models in your own environment
- Evaluation, cost control and governance aligned with GDPR and ISO 27001
Master Data Management
Governed, shared master data for customers, suppliers, products and finance – that the whole business can trust.
- MDM strategy, data model and golden record
- Data governance: ownership, roles, processes and decision rights
- Data quality rules, measurement and continuous monitoring
- AI-assisted matching, de-duplication and enrichment
Data Migration
The full migration journey to SAP, Dynamics 365 or Infor M3 – from assessment to cutover and hypercare.
- Strategy, planning, waves and safe cutover with fallback
- Source and target models, field mapping and transformation rules
- ETL in Python with automated validation and reconciliation
- AI-assisted mapping, profiling and validation
AI Engineering in practice: Read about imfund, where I run a whole team of AI agents →
Data migration – a method in eight steps
Every data object goes through the same eight steps, built on PDM v3 and refined over the years. That makes the migration repeatable, measurable and easy to track.
- 1
Identify scope
What data should move?
- 2
Understand the models
Source and target.
- 3
Map the fields
Source to target.
- 4
Build the ETL
Extract, transform, load (Python).
- 5
Cleanse data
In source, transform or after go-live.
- 6
Test migration
Key people and super users verify in DEV.
- 7
Broader test
More of the business validates in TEST.
- 8
Cutover
Plan go-live with fall-back.
Steps 6 and 7 – the test migrations – are owned by the business. Read how I work in practice →
AI on the client's terms. Depending on the organization's requirements, AI solutions can be built using either leading commercial models accessed through APIs or open-source, self-hosted models running within the customer's own environment. The focus is always on secure data handling, transparency, and compliance with GDPR as well as the customer's internal security policies.
Approach
I bridge business and technology — ensuring clear ownership, high data quality and sustainable governance.
As the consultant behind BBIMP, I combine deep technical understanding with strong business insight, from strategy and data modelling to validation, load and cutover. Pragmatic, transparent and proven.
You work directly with me throughout the engagement — no handover to junior resources, just senior expertise from the first meeting to hypercare.
About me
I'm Isak La Fleur Engdahl – the consultant behind BBIMP. For over ten years I've led data migrations and master data programmes across mining, manufacturing and retail: global Infor M3 and IBM InfoSphere rollouts at Atlas Copco and Epiroc, and the data migration at one of Sweden's largest retailers. Today I also build AI agents in production – for example in the imfund platform. With an MSc in Industrial Engineering and Management, I pair deep hands-on Python and SQL with the business judgement and communication that large migration programmes demand. Hire BBIMP and you work directly with me – no handover to junior resources.
ERP & MDM experience
Platforms I've worked with in live migration and MDM programmes.
ERP
- SAP S/4HANA
- Infor M3
- Dynamics 365 F&O
- Dynamics 365 Business Central
- BPCS
- MFG/PRO
MDM
- Informatica MDM
- IBM InfoSphere MDM
Data & ETL
- Python
- SQL
- Apache Spark
- Microsoft Fabric
- Databricks
- Snowflake
- dbt
- Azure Data Factory
- Apache Airflow
- DuckDB
- Polars
- pandas
- Parquet
- Streamlit
AI engineering
The stack I build production-grade AI with – from model APIs and agent frameworks to self-hosted models and evaluation.
Models & platforms
- Anthropic Claude API
- OpenAI API
- Azure AI Foundry
- Amazon Bedrock
- Google Vertex AI
Agents & orchestration
- Model Context Protocol (MCP)
- Claude Agent SDK
- OpenAI Agents SDK
- LangGraph
- Paperclip
- Hermes Agent
Self-hosted & retrieval
- vLLM
- Ollama
- Hugging Face
- pgvector
Engineering & evaluation
- Claude Code
- OpenAI Codex
- Cursor
- Langfuse
- Promptfoo
From the blog
Thoughts and lessons on AI engineering, master data management and data migration.
A five-minute check-in on my AI agent company – from the football sideline
imfund's AI agents now work on their own on a Linux server at home, and I can get a status update from the Chief of Staff via Telegram. On Saturday that meant one quick check-in on my phone during my daughter's football match – the rest of the time I watched the game. They won 5–2.
Read more
imfund: two people and an entire company of AI agents
At imfund, two people sit on the board while 16 AI agents do the work – each with a manager, a job description, a schedule and a budget. Here's how it's organised, which decisions stay with the humans, and why it's more about governance than about AI models.
Read more
The data object migration plan: what moves, and when
A migration plan isn't a field-mapping spreadsheet — it's a dated runbook of what data moves into the new system and on which day around go-live. Bulk loads versus delta loads, why master data and transactions freeze on different dates, and why the cleanest way to migrate the hardest objects is to clear them in the real world first.
Read moreContact
Want to put AI agents to work, get your master data in order or facing a migration? Get in touch.