Hello there 👋

Field full of buttercups

Writing this from somewhere in the west of Gloucestershire, where I live with my partner and our two dogs, Letty (Irish setter, rescued from Greece, loves sleep) and Cassie (border collie, loves chaos).

Most of my working life has been at Holiday Extras, where I am generally interested in making data work, nominally as a data scientist which means sitting somewhere between development, analyst and architect.

What that really means is I spent lots of time bouncing between figuring out what we need to do with data to help the business, what we need to build, buy or adopt, to make that possible … and then actually getting on with implementing it.

Man at desk with head in his arms

Getting there

So, while I was writing up (code for: funding ran out) my PhD in Medical Image Processing, and back before everything was python libraries, I started with writing screen scrapers to check our competitor’s prices. This proved useful! Then came importing that data into our then brand new, sparkly data warehouse and quickly moved on to debugging and building various bits of ETL pipeline for our in-house web tracking (around the same time as urchin.js was a thing but with more detail). With a brief detour running CubeViews on WebShere, quickly corrected, then split testing, lots and lots of reporting, debugging web pages, and finding out where stuff was going wrong on the web-site.

Recently, I have spent more time getting bits of Machine Learning running getting our website a little better, helping with product navigation and generally trying to make the
user experience a little better. Again, lots of this is really about helping leadership work out how (if!) we should actually use ML

Along the way, I have seen lots repeated patterns, so the architecture hat comes on in getting tools like Looker, Airflow, our Schema Registry, Big Query and ML-Engine / AI-Platform / Vertex AI adopted across the business.

When I say architecting, it’s really the less glamorous process of building out lots of proof-of-concepts (or proof-of-failures) working out exactly where things break for us down the road. The least broken wins.

Interesting things at time of writing

Current tech I am keeping my eyes on: dbt, metric stores like Transform, feature stores and generally trying to keep my head wrapped around the explosion in new data tools and platforms, mostly with an eye on trying to work out what might be useful, and what will be noise.

Daily tools - Python, SQL and whichever frameworks we are supporting now. Big fan of Google’s Cloud Platform, which makes lots of infrastructure easy enough I seem to be able to use them.

Outside work, more on the mountain bike than the windsurfer now I have moved from the coast, or DIY-ing various bits of our cottage and generally losing to nature in the garden. Occasionally exploring vegan options in my new Ooni.

You can find me on lurking on Twitter and Medium, which are great places for getting to know the data landscape.

Man walking with dog in woodland