Hi, I'm Erik.
I build data pipelines, dimensional models, and Power BI reporting for industrial clients.
Eight years of hands-on data work at IVL Swedish Environmental Research Institute (2018–2026) for large manufacturing and automotive clients: refreshable pipelines in Power Query, Bill-of-Materials data shaped into dimensional models, and Power BI dashboards used by procurement, risk, and strategy teams. Python for analysis and optimization across research and consultancy projects.
What the work looks like
The work usually starts where a spreadsheet stopped scaling.
Pipelines with provenance and completeness checks built in
Layered pipelines that turn fragmented source data into reports where every number traces back to its source.
Dashboards the business actually opens
Interactive Power BI reporting built from stakeholder interviews, so procurement, risk, and strategy teams can answer their own questions.
Refreshable, not rebuilt every quarter
Pipelines with typed functions and config-driven parameters, so the next reporting cycle is a refresh rather than a rebuild.
Where I go deep
Bill-of-Materials data as a dimensional model
A 400,000-row bill of materials modeled as a fact table with material and classification dimensions, so critical-raw-material usage could be sliced by part, supplier, and region.
Regulation encoded as a layered pipeline
The EU Battery Regulation's Circular Footprint Formula built as ~40 Power Query queries in bronze, silver, and gold layers, with typed M functions centralizing the regulation's sign conventions.
Medallion lakehouse to semantic model
An end-to-end Microsoft Fabric platform: medallion lakehouse, dimensional fact and dimension tables in the gold layer, and a Power BI semantic model in DirectLake.
Transformation logic pulled out of notebooks and tested
PySpark transformations extracted into functions specifically so they could be unit-tested by pytest in GitHub Actions on every push.
Background & Credentials
Experience
- Eight years at IVL Swedish Environmental Research Institute (2018–2026)
- EU-funded research projects (ELiMINATE, SIREN)
- Multi-year engagement for a Fortune 500 automotive manufacturer, under NDA, via IVL
- Contributed to 40+ research and consultancy projects as analyst, LCA practitioner, and project leader
Certifications
- Microsoft: Fabric Data Engineer Associate (DP-700)
- Microsoft: Fabric Analytics Engineer Associate (DP-600)
- Microsoft: Azure AI Fundamentals (AI-900)
- Microsoft: Power BI Data Analyst Associate (PL-300)
- Microsoft: Azure Fundamentals (AZ-900)
- SQL for Data Science — University of California, Davis
- Google Data Analytics Certificate — Google
Technical Stack
- Data: SQL, Python, Power Query / M, PySpark, Microsoft Fabric, Power BI, Azure SQL Database
- Practices: Dimensional modeling (Kimball), medallion architecture, ETL/ELT, Git, unit tests in GitHub Actions
- Analysis: Lifecycle Assessment, Optimization Modeling
- Domain: Critical Material Risk, EU Battery Regulation, REACH
Research & Publications
Academic research and reports from collaborative industry projects.
Optimizing Resource Recovery of Lithium-Ion Battery Recycling
Material flow analysis and reverse-logistics network optimization for end-of-life lithium-ion batteries in Europe up to 2030, comparing recovery technologies on cost and CO2 (C801 Report).
Plastics in Passenger Cars
Analysis of plastic materials usage in passenger vehicles and implications for recycling and circular economy (C454 Report).
Lithium-Ion Vehicle Battery Production
Meta-analysis of life-cycle energy use and greenhouse-gas emissions from lithium-ion battery manufacturing, revising the earlier 150-200 kg CO2-eq/kWh estimate down to 61-106 kg CO2-eq/kWh on the basis of commercial-scale production data (C444 Report).
My Approach
Start with the end report
Design data systems backwards from what stakeholders actually need to see and decide.
Automate what repeats
Identify manual processes that consume time each cycle and build reliable automation.
Build for audit trails
Every number traceable back to its source — which is what makes a figure defensible when someone asks where it came from.
Document for the next person
Create systems that others can maintain and extend, not black boxes only you understand.
When not building data pipelines, you'll find me hiking the Swedish west coast, reading about energy systems, or experimenting with new visualization techniques. The work I like best sits where technical rigor meets genuine curiosity about how a system actually behaves.
Interested in connecting?
I'm always happy to discuss data engineering and BI work, research collaboration, or just exchange ideas on modeling messy industrial data.
Get in touch