The first thing that comes to mind is working with data at scale. There are fundamental differences between personal projects and real-world company datasets - one query can return hundreds of thousands of rows of data. I'm proud to have sharpened my data cleaning and preprocessing skills, whilst strengthening my Python and SQL data language abilities.

I also learnt how an international team can operate successfully - my team spans across Budapest, London, and Tallinn. Despite being spread across multiple locations, our productivity and collaboration has been excellent.

Another valuable lesson was the importance of having exact, objective definitions for everything in analyst work. Aligning on definitions and a source of truth can be tricky, which underscores the importance of good communication skills.

Perhaps the most important lesson from my first week was understanding the crucial need to deeply comprehend (and question) the product you're working on. What drives excellence is a deep understanding of the product's flows, the company's mission and vision, and the courage to ask whether there's a need to be doing what the team is currently doing.

One of the most valuable learning experiences also came from receiving thoughtful, sometimes challenging feedback from teammates. This taught me how to critically self-review my own code multiple times before submitting it, ensuring it's the best version possible.

A significant turning point was the mid-internship review with my lead. The constructive feedback helped me pinpoint exactly what I needed to work on. For example, I learnt to prioritise tasks more effectively rather than trying to do everything in parallel. A more focussed, sequential approach served me much better in the second half of the internship.

Early in my internship, I struggled to understand why the results of a query I wrote returned very counterintuitive results from the database – a very typical and common challenge for an analyst. I couldn't pinpoint the issue in my code and spent a lot of time thinking about what might be going wrong, forming hypotheses one after the other, but never actually testing them. 

I brought this issue up in a 1:1 with my lead, and they immediately started probing the root cause, showing me what good hands-on analysis looked like. The issue turned out to be with a certain field in a table that I was counting on to be present, but actually stopped getting populated at some point earlier this year. 

My lead’s explanation was incredibly helpful, and this method of conducting analyses has stuck with me. I learn a lot by seeing how other people work, and having my lead troubleshoot this issue with me firsthand, showed me that I could, and should, increase the rate at which I form and validate/invalidate hypotheses - a big part of an analyst's job.