About

Ian Ozsvald picture

This is Ian Ozsvald's blog (@IanOzsvald), I'm an entrepreneurial geek, a Data Science/ML/NLP/AI consultant, author of O'Reilly's High Performance Python book, co-organiser of PyDataLondon, a Pythonista, co-founder of ShowMeDo and also a Londoner. Here's a little more about me.

High Performance Python book with O'Reilly

View Ian Ozsvald's profile on LinkedIn

ModelInsight Data Science Consultancy London Protecting your bits. Open Rights Group

Building Python Data Science Products

I have a feeling that I’ll write a book on how to Build and Deliver Data Science Products using Python. This will be based off of 15 years commercial experience building successful data science products for companies, speaking and teaching on the subject for 6 years at international conferences and authoring two other successful books (including O’Reillys’ very-well-reviewed High Performance Python). I have a draft set of notes online and I’m very interested in your feedback.

Professionally I also coach data science teams as an interim senior data scientist. My coaching experiences will define some of the book’s sections with a focus on problems that are known to occur in existing data science teams and safe routes through them to successful deliveries:

“Ian coached our team when we needed some extra technical firepower, and provided that in spades. He slipped into a role providing technical leadership to a new bunch of people, and energised every project to which he contributed. He also straightened our path towards best practice, with a combination of good sense and business experience, for which generations of my team will be grateful.” – Alice Jacques, Channel 4

By joining my list (below) you’ll hear about my early development and you can help me design what I’ll write, so you’ll get the book you need. Topics I plan to cover include:

  1. Quickly getting, understanding and visualising data
  2. Rapidly moving to use reliable data (where 80% of your time might be wasted!) with efficient techniques to clean, normalise and prepare your data for standard industrial use-cases
  3. Building early models to show your team that you’re on the right track and to increase management but-in
  4. Validating that your models are working correctly and will generalise to the real-world (so you don’t get blind-sided when you deploy them and they don’t deliver like they did during testing!)
  5. Successful approaches to deploying Python projects using testable modules, logging and microservices to simplify your devops life
  6. Approaches to debugging models and identifying when your data is making the algorithm go wrong

I’ll inform you when I’m working on the book when you join my list:

* indicates required