Project volunteer Tom Reeve

“Even for a ‘small’ project without the technical depth you might see in a full commercial project, it was possible to have a very real impact on a real problem in a community.”

Pronouns He/Him
Roles Director of Cloud Technology, Intent HQ | Data project volunteer
Links https://github.com/TRReeve

How did you first get into data?

I was always interested in processing and pulling insight out of the world around us. I got my start as a business insights analyst after studying political analysis and finding it wasn’t exactly what I was interested in, and was able to do so thanks to taking some modules in quantitative methods and economics.

Ever since that I have kept discovering more powerful layers sitting underneath my spreadsheets, from SQL to Python to the underlying query engines, and operating systems and chip architectures themselves. Being able to work with that data is a more powerful means of insight that allows for practical problem solving, so despite the start in a different field, I’m very happy that things worked out in this way.

What do you wish you’d known when you started to get into data?

Before getting into technical solutions and trouble-shooting, it’s a good idea to take time to actually understand what the problem you are trying to solve is, whether working on data or software projects. That rule applies to what a tiny NGO does just as much as it applies to a billion-plus-turnover technology company.

What was your first volunteering experience with DataKind UK like?

It was fast. I worked with a very talented data visualisation journalist on a backend with PostGIS that would render polygons on a map that ran entirely locally and could be customised to analyse access to recycling centres. l discovered how much of a learning curve there can be when you are not using your ‘normal’ tools from your day-to-day job.

Setting up simple tooling from zero can take up surprising amounts of time. As someone used to working in a software team where I had built out cloud infrastructure, dev-environments, and internal libraries around our problem space, it was quite a shock to go back to zero. Tools like docker came to the rescue massively. I was pretty happy with the potential of the solution, but it was amazing how quickly the end of the volunteer event comes.

I think what hooked me and made me want to come back was that even for a ‘small’ project without the technical depth you might see in a full commercial project, it was possible to have a very real impact on a real problem in a community. And this is thanks to some very talented, dedicated people who do this full time, and were able to leverage that technical knowledge.

What are the most important things you learned from volunteering?

As an event volunteer, and then as a project volunteer, the most important thing I learned is that it is a lot easier to have success with an incredibly simple use case, and iterate it to demonstrate the art of the possible, rather than aim to build the greatest possible solution in a weekend (this approach also works very well for a lot of commercial work). This also means a wider variety of skill sets can work on the problem, which normally gives you interesting outcomes.

What is a data project that inspires you?

Both as an end user and as someone who is interested in the guts of query and storage systems, DuckDB is a marvel for getting so much right at the technical level and operating at the cutting edge of query execution and storage, while also nailing a good local user experience without forcing you down the ‘big cloud service’ route as a default.

It’s so quick for picking up, preparing, and managing data. I’ve used it for every project to quickly prepare data seamlessly between SQL and embedding into software scripts. It’s really well documented, really easy to use and covers the ‘data exploration’ as well as, if not better than, the ubiquitous Python Pandas, while also playing nicely alongside it and many other notable dataframe tools.

Is there a resource or tool you’d recommend?

It’s very old fashioned, but if you want to learn something, reading a book on it by someone who knows the topic deeply both gives you the direct information on ‘how’ in much more depth, and also gives you a lot of pointers into the related topics that you don’t know you needed to know. This is similar to how someone interested in LLMs might not be exposed to how they link to Markov Chains, and in turn how they can be applied to random walk problems in geospatial domains.

In online stuff I would recommend the Pragmatic Engineer newsletter on Substack. It’s a very well-written newsletter with great deep dives into wider technology industry trends. I also follow Jesse Anderson.

Tell us something completely non-data-related about yourself!

My trash TV pick is Real Housewives of Beverly Hills.