Over the past 12 months, Kainos’ Data and AI practice has grown significantly as we continue to support our customers on their maturity journey from tactical, to strategic and transformational use of data and artificial intelligence.
To support our continued growth, we have hired a new Head of Data Science, Alexey Drozdetskiy.
We had the pleasure of interviewing Alexey as he shares insights on his lengthy career in academia, why he transitioned into a role in industry and his plans for supporting our customers to realise genuine business benefits from operationalised artificial intelligence.
What are your main priorities as Head of Data Science at Kainos?
My first priority in my role is to demonstrate to our customers (existing and new) how our Data and AI Practice can help them achieve their goals in the most efficient, reliable, and optimal way.
I want to showcase how we apply our best practice approach to designing and delivering solutions and how we use our comprehensive MLOps approach, and scientific rigour to take ML-solutions from experimentation through to production. I want to help our customers explore solutions and enable them with the knowledge and skills required to support their solutions in the future.
My second priority is to continue building out our Data and AI team, ensuring they have the right skills to serve our customers’ needs in the best way possible. The projects Kainos deliver, positively affect the lives of millions of people – and they are delivered thanks to the brilliant people that work here – fostering their talent is critical to our continued success.
Tell us a bit about your background:
Before joining Kainos I worked for five years at Altius (and DataSparQ), where I successfully led on the delivery of multiple, complex AI/Machine Learning (ML) driven solutions from, running advisory workshops to implementing production services across almost every industry vertical.
Before that, I worked in the academia for almost 20 years. The most important part of that was working at the European Council for Nuclear Research (CERN) on one of the two major experiments on the Large Hadron Collider. I focused on the Higgs Boson (Nobel Prize) discovery. And for a couple of years after CERN I worked in Computational Biology, using Neural Networks to predict protein structures.
What made you move from academia into Data Science in industry?
There were several reasons. I suppose the most relevant and most interesting reason for this post is: however interesting and challenging the world of academic research was – after working there for multiple years, I wanted to explore alternatives. Data Science was an obvious choice – it is challenging, dynamic, always keeps you on your toes – the complexity of questions, complexity of human-generated data in businesses or public sector organisations – is unparallel and rivals the most complex scientific setups. I never looked back after switching to industry.
What do you get up to in your spare time?
Continuous learning, education and experimenting with ideas form a big part of my spare time. The field of Data Science is vast and continuously expanding in the breadth of coverage and sophistication – there are always new topics to explore, which keeps me busy.
Plus, of course, lots of reading and some training to keep healthy. Nothing too crazy. Though, I’ve done a few 100-mile trail runs and I have been practicing yoga for about 17 years.
How can our customers leverage AI in their businesses? What are the biggest opportunities our customers can leverage by using AI?
Most organisations have lots of passive, or not optimally used data; from documents that are not digitised or structured, to multi-dimensional data that is ignored in favour of outdated simplistic “decision tree”-style approaches, or insights from 1-2 dimensional visualisations done by hand.
Business critical KPIs could be dramatically improved by comprehensively utilising business data, combined with public, open-source data. First, this could be demonstrated through an initial PoC/MVP running in parallel with the current practices. Second, it has to be realised by productionising the AI/ML solution; taking the models from experimentation in silos to production and making AI a functional, repeatable, expandable component of a business.
Machine learning, data science and AI in general are becoming more and more commonplace in all kinds of organisations; public and private sectors. However, the depth of experience, especially around operationalising machine learning, is often lacking. We can help customers close this capability gap.
If an AI solution is not operationalised and not used – it may as well be non-existent. However, that simplye isn’t an option for organisations wanting to retain a competitive edge – as it’s only be a matter of time for other competitors to get there.
I am always keen to speak with customers to discuss their data strategy and help operationalise their data. Whatever stage you are at in your journey, feel free to drop the Data and AI Practice a message at email@example.com and we can see how we can help.
Are you interested in finding out how we have operationalised Data and AI solutions for our customers? Check out our Machine Learning Operations (MLOps) services page to read our latest case study or speak to an expert.