Engineering managerData platformsFinanceApplied AI
Engineering manager building finance data and AI products
I lead a team of data and BI engineers building finance and risk data products, and I look for the places where applied AI makes finance work simpler. Over about 28 years I have moved from hands-on engineering to leading delivery in finance, telco, energy and deep tech.
Focus
Building data platforms and finance data products that people trust, today on Databricks and Power BI, before that on Azure for group tax reporting.
Bringing AI into core finance processes where it genuinely helps the people doing the work, not as a showcase.
Leading engineering teams and complex programmes, from deep-tech R&D and national integrations to platform rebuilds.
I'm an IT leader with a passion for technology and data-driven problem-solving. For about 28 years I have moved from hands-on engineering to leading teams through complex projects, integrations and process improvements.
I started as a systems engineer with Schlumberger in Norway, then worked in solution architecture, professional services and programme management at Oracle, Sun Microsystems, IBM and Slovak Telekom. Since then I have led delivery in deep tech, robotics and finance, and today I lead data and BI engineering in finance.
Most hard problems I have worked on were not purely technical. They needed cross-functional teams that understand each other, a clear plan and data people can trust. Helping put that together is the part of the job I enjoy most.
Experience
Previously Aptitude IT Team Lead, Aug 2024 – Oct 2025
Outcome: Lead a team of 17 data and BI engineers building finance and risk data products on Databricks and Power BI, and drive applied AI in core finance processes.
A data and BI engineering team serving finance and risk. My part is direction, delivery and people, and finding where applied AI can take work out of core finance processes.
Outcome: Led the delivery of a complete rebuild of Swiss Re's group tax calculation and reporting platform on Azure, covering IFRS and US GAAP.
A contract engagement through Aardwark. The group's tax calculation and reporting platform was rebuilt end to end on Azure, with reporting under both IFRS and US GAAP.
Outcome: Delivered a vision-guided robotics solution for warehouse logistics with Photoneo, now part of Zebra Technologies, directing AGVs that moved goods between shelves and picking stations.
Warehouse logistics where machine vision guides automated guided vehicles (AGVs), delivered together with Photoneo.
Outcome: Directed the R&D programme behind a breakthrough plasma drilling technology, leading more than 40 scientists and engineers.
Deep-tech research and development in plasma drilling, run as a programme with a large team of scientists and engineers.
Outcome: Led the team behind a cloud platform for self-configuring WiFi routers, an early step towards intelligent, automated network management.
A software team building a cloud platform that let WiFi routers configure themselves.
Outcome: Led a national programme for the Ministry of Transport, integrating traffic, mapping and telemetry data from mobile operators, road authorities and the police into a single application that went live nationwide.
A public-sector programme for the Ministry of Transport that brought data from several organisations together into one application.
Projects
Personal side projects in AI and data, built outside my day job.
AI assistant that reads a bank statement and finds the matching invoice for every payment in your inbox.
Debit Detective turns a PDF bank statement into a clean list of outgoing payments, then hunts down the receipts. Gemini reads the statement, extracts each debit with its date, amount and a cleaned-up merchant name, and returns structured JSON. The app then searches Gmail for each payment by amount, merchant and a ±5-day window, and labels every matching email as INVOICE, so month-end expense claims or bookkeeping take minutes instead of an evening. Results can be exported to CSV.
View on GitHub: Debit DetectiveA daily Databricks pipeline that turns my running data from Runalyze into weekly training insights.
A small but production-style data platform for my own training. Every night a Databricks job pulls new activities from the Runalyze API and runs them through a medallion architecture: raw JSON lands in a Bronze Delta table, gets parsed, typed and de-duplicated into Silver with MERGE upserts, and is aggregated into Gold tables for weekly mileage, weekly pace and time spent in heart-rate zones. Secrets live in Databricks secret scopes, and the whole job is defined as a Databricks Asset Bundle, deployed automatically by GitHub Actions on every push to main.
View on GitHub: Runalyze on DatabricksBackground
Recommendations
[TODO: Paste a real recommendation from LinkedIn here, with permission.]
[TODO: Paste a real recommendation from LinkedIn here, with permission.]
[TODO: Paste a real recommendation from LinkedIn here, with permission.]
[TODO: Paste a real recommendation from LinkedIn here, with permission.]
Notes
A short post about once a quarter on data, AI and delivery.
Contact
I usually reply within a few business days.