Path to power
I have had a varied career over the last 12 years, which has helped me in the wide variety of things I do as an independent consultant, author, speaker and chief data scientist. My PhD thesis investigated statistical methods and designed experiments for tuning AI algorithms. This has proven important in recent years as the prevalence of algorithms has led to industry concerns about ethical, explainable and reproducible decision-making.
My pre-sales consulting with a data-mining fraud solution really helped me appreciate fast-paced delivery, the complexity of data products and the importance of understanding a business domain. My years leading forensics analytics teams in “big four” accountancy firms opened my eyes to managing delivery risk and communicating complex analytics at the right level for senior stakeholders.
I hope my book “Guerrilla Analytics” gives managers and practitioners starting out the principles and practices that will help them accelerate their careers. At Sainsbury’s, I have put all of this into practice by building the retailer a capability that modernised its ability to make complex retail decisions at massive scale.
What is the proudest achievement of your career to date?
Writing Guerrilla Analytics has certainly opened doors for me. A book is a huge investment of time, but it helped clarify my thoughts on the best operating principles and practices for analytics and data science, as well as growing my network internationally. I’ve been able to use that experience in my consulting work, my speaking engagements and to raise standards in the data strategy and design principles for algorithms and data at Sainsbury’s.
Who is your role model or the person you look to for inspiration?
I always look up to those who are great communicators and who manage to transition from the rigour of academia to the pragmatic budget and risk focus of industry and consulting. Peter Norvig and Cassie Kozyrkov of Google have certainly achieved this. I also admire the hundreds of engineers and scientists who are making incredible advances in cloud technology and algorithms. There are too many people to mention – I must follow at least 50 blogs to keep up to date.
Did 2019 turn out the way you expected? If not, in what ways was it different?
Yes. My aim in 2019 was to build on the previous year’s success with my specialist cross functional team of engineers, scientists and product owners. I made some key hires, grew our algorithm capability from a principles and practices perspective and transitioned the capability into traditional engineering teams in a new technology function.
What do you expect 2020 to be like for the data and analytics industry?
The convergence of data science, engineering and cloud will continue. The more advanced organisations are recognising that cross-functional teams are essential to deliver pragmatic algorithms that give a return on investment. For the industry, that means a continued focus on skills, organisation design and hiring the people with the soft skills to work in cross functional teams.
Data and technology are changing business, the economy and society – what do you see as the biggest opportunity emerging from this?
Algorithms and the data that feeds them are changing our lives like never before. Boring and low value tasks can be automated with confidence. Sophisticated decisions that are too complex to be made by a human can now be made by machines – think complex decisions with many competing objectives and constraints. This new decision-making is an opportunity for us all to create efficiencies for our businesses and society, and to create new products that delight our customers.
What is the biggest tech challenge you face in ensuring data is at the heart of your digital transformation strategy?
The biggest tech challenge from a data science and machine learning perspective is choosing the right technology with the right priority and focus in a data transformation programme. Many enterprises will stick to what they know – effectively building old fashioned, highly structured warehouses in the cloud at considerable expense. These warehouses, of course, have value for analytics and BI but algorithms have different data requirements – deeper histories, latency, streaming, point-in-time history. Knowing the technologies to choose so that all data consumers can add value is the biggest challenge.