I'm a data scientist working at the intersection of technology and design. Reformed astrophysicist & former e-Research/data consultant.
Stitch Fix's Machine Learning Algorithms

Stitch Fix's Machine Learning Algorithms

After exhausting all available episodes from Ologies, The Design Files Talks, Data Stories, and IDEO’s Creative Confidence podcast, I finally moved onto Guy Raz’s How I Built This podcast, starting with the interview with Katrina Lake from Stitch Fix. I first knew about Stitch Fix when ex-astronomer turned data scientist Eli Bressert joined the company back in 2014 as head of their research and development. I also became aware of their growing data science team when I was researching the Insight Data Science Fellowship, specifically the research backgrounds of fellows and the companies they moved to after the program. At the time I hadn’t fully appreciated how data-driven the company was, nor did I know much about Katrina Lake and her remarkable story. Suffice to say the interview wi really worth listening to.

So naturally I spent the following morning looking through the Stitch Fix algorithms blog, listening to a second interview (from Data Minds) with Eli Bressert about his time at Stitch Fix and how the R&D and product teams evolved during his time there, and reading through their wonderful visual Algorithm’s Tour. This is such a fantastic resource for understanding why, when, and how Stitch Fix’s implement specific algorithms for specific purposes. It’s a level of transparency and shared knowledge that you rarely see from tech companies and it’s such a great way to find out what might a typical day look like if you were to join the team. At some point in the not too distant future I want to visualise tech companies according to their the most common algorithms – recommendation, search and discovery, object recognition, natural language processing etc., or purpose, and if possible drill down into specific algorithms along throughout the product journey.

While looking at the current algorithms team I was pleased to see two more ex-astronomers (research-based) have joined the team, Dave Speigl (Merch Algorithms), and Mark Dijkstra (Global Optimisation), as well as Ada Draginda (Data Engineering), a former astronomy software developer at CFHT.


Designing AI Systems

Designing AI Systems

Best practises for combing data science and design thinking

Best practises for combing data science and design thinking