HOW I WORK
Working with AI
I use generative AI every day to automate routine steps, draft and check work faster, and keep learning as new tools arrive. This page explains how, including how this website was made.
This website
HOW IT WAS MADE
This website was built with the help of AI. I used large language models, mainly Claude and ChatGPT, to structure the pages, draft and proofread the English copy, develop the layout and write the code for the site and its article pages. Through APIs, I connected these tools to Composio, Thunderbit and Zernio, so they worked directly with the underlying data.
The content is my own. Every role, date and figure comes from my CV and my project work. I reviewed every page before it went live. Nothing was invented, exaggerated or rounded up by AI.
Continuous learning
KEEPING UP
The number of AI tools is growing quickly, and so is the demand for people who can use them well inside a project. I treat AI as a tool for continuous learning: I test new tools on my own work first, keep what saves time without lowering quality, and drop the rest.
In 2026 I completed the certification Generative AI for Business Consultants. I write about practical uses in the Articles section under the tag Generative AI.
Daily work
WHERE I USE IT
In my daily work I use AI to automate processes and to take the routine out of analysis, documentation and reporting. Much of it is code: AI helps me write the scripts behind recurring reports, first in MS VBA, now also in PowerShell.
• Portfolio reconciliation (a Leading Swiss Bank, 2025). I used AI to write the MS VBA code for the daily reconciliation reports and overviews. Complex logic was ready to test significantly faster than coding it by hand.
• Application decommissioning (a Leading Swiss Bank, 2026). For the decommissioning of 17 legacy applications, I used AI to write the PowerShell scripts that produce the reconciliation reports and overviews. Same effect: complex logic, coded significantly faster.
Every script was tested against the source data before it was used.
Division of labour
WHO DOES WHAT
The tool
• First drafts and structure
• Summaries of long material
• Scripts and formulas for repetitive steps
• Proofreading and consistency checks
Me
• The question and the context
• Checking every output against the source
• Decisions and recommendations
• Accountability for what I hand over
Client data
CONFIDENTIALITY
In a client project, the client's rules come first. I use only the AI tools the client has approved, and I do not enter client, portfolio or personal data into public AI tools. Banking secrecy, the GDPR and the client's information security policy apply to AI as they do to any other tool.
Planning AI use in your project?
I can join your team as a Business Analyst who already works with these tools, within your rules.