AI that supports
real operations.
As AI Innovation Trainee at BMW Financial Services Northern Europe, I built and launched three internal AI applications for Operations teams in Sweden, Norway and Denmark.
- Type
- AI + Operations
- Status
- Experience
- Period
- Apr – Sep 2025
- Tools & technology
- Retrieval-augmented generation, LLMs, Workflow automation
01Context
Operations teams in financial services work with a lot of internal knowledge: processes, guidelines and routines spread across documents and three countries. Finding the right answer quickly matters, and so does the answer being aligned with internal guidelines.
02My role
AI Innovation Trainee. I built and launched three internal AI applications and worked directly with Nordic Operations stakeholders to find recurring challenges worth solving.
03Process
- Working with Operations stakeholders in Sweden, Norway and Denmark to map recurring questions and bottlenecks.
- Developing document-grounded retrieval workflows so answers come from internal guidelines, not guesses.
- Building, launching and iterating on the applications with the people who use them.
04Outcome
Three internal AI applications launched for Operations teams across the Nordics, and practical experience of bringing AI into a regulated environment where trust and accuracy come first.
05What I learned
- Useful AI starts with the workflow, not the model.
- In operations, a correct “I don’t know” beats a confident guess.
- Security and governance shape the design from day one.