Data Science Consulting: Generalist Path

Preparing for the first step in your consulting journey

Publish date:

Article by Dominik Deja – Data Science Manager at Capgemini Invent

In the last post, we discussed what data science consulting is all about and provided explanations for the most frequently used terms and buzzwords. Today, let’s deep dive into the data science consulting itself. We’ll explore what it means to be a data science consultant by giving a concise summary of the role and by analyzing 3 different use cases, from owning a selected piece of work, through delivering an entire Proof-of-Concept, up to leading an AI transformation.

Coming back to our definition, Data Science Consulting is a practice of helping organizations to improve their performance by applying scientific methods, processes, algorithms, and systems focused on using data to allow for automated, data-driven insights generation and decision making.

How does “Generalist” relate to this definition? As long as you’re not focused on any particular set of techniques or business domains, and you’re keeping your skills and attention sharp, but relatable to wide range of businesses – you’re a generalist. If you start to specialize in, for example, developing risk scoring models for banking, then you’ll naturally start to specialize in delivering data science solutions for banking industry. Both paths are equal in terms of opportunities  – it’s more of a personal preference and the choices you’re making along the way.

How does being a data science consultant vary from being a data scientist itself? At first glance, it’s indiscernible – most of data scientists are concerned with mathematical models carved in Python code and kept on a secure repository from which it is continuously integrated, tested, and deployed in development/stage/production environment. Sometimes, the models will be different, the code will be written in some other programming language, or the deployment pipeline will vary, but in principle, it’s the same set of tools.

What differentiates those two roles are the practices of their everyday work – while a pure data scientist will, for most of their time, write code and attend dailies, a consultant, on the top of the standard data science work, will prepare client-ready reports based on the work they did, attend client calls, explain directly to the client what has been done, and gather the requirements. In other words, a data science consultant will merge the client-facing part of consultant responsibilities with the code-writing part of data science responsibilities. A one-man-army.

Of course, it takes time to become such a one-man-army, and no one requires fresh-graduates, or people who recently requalified themselves into data science, to cover the full range of responsibilities and duties of a data science consultant. Usually, one grows into this role – starting from supporting a selected, narrow part of a bigger project, through being responsible for a selected piece, coaching and mentoring younger colleagues, and finally being responsible for delivering the whole data science scope, or even a project (in case of small-enough ones). As you become an expert, usually, you’ll have to choose whether you’re more into the technical aspects of your work or prefer the people-related topics. At a cost of oversimplifying, big tech companies prefer most of their experts becoming technical experts, while in the consulting, people-focus is preferred.

So, let’s say that you just started your journey as a junior data science consultant. What to expect? Usually, you’ll be offered to join a project, which is already ongoing, where there’s some technical, quite well-defined work to do, and where there’re already some data scientists in. Your exposure to the client will be limited, and you’ll support the work of some more senior data scientists who’re already on the project. They’ll coach and guide you, assign more or less challenging technical tasks and expect you to deliver those on time. You might be asked to support reports preparation for the client, but it all depends on the project – in most cases, you’ll focus on delivering well-defined tasks, relatively small puzzles building the bigger picture.

As you’ll grow, you’ll become more and more independent of others – meaning that you’ll have more freedom to express yourself in your work, but also meaning that you’ll have to take the responsibility of defining your work. You’ll start managing your own streams, and lead, first interns, then juniors, and as you become a senior, even mid data scientists. With time, also your client exposure will increase – from taking part in client calls and supporting your manager in case of tough technical questions, through preparing and presenting the reports you’ve created. The more you’ll grow, the more independent you’ll become. Eventually, you’ll be able to deliver small projects end-to-end yourself, or lead and deliver specific workstreams, while interacting with the client.

Being a proficient, senior data science consultant, you’ll face a choice – to focus on technical aspects and becoming a so-called content expert, or to follow a more people-oriented path, where, you’ll spend most of your time with your clients and manage the teams. It’s fully up to you, which path suits you the best. As a technical expert, you’ll shape the solutions offered to the client and make sure that they’ll work as expected. You’ll join calls if deep expertise is needed, and you’ll be widely recognized by your knowledge. As a people expert, you’ll make sure that your team of data scientists is working well and that clients are willing to cooperate with your company. You’ll get involved in bigger digitalization/AI transformation programms, in whichyour understanding of how to win people will be equally important as your technical expertise. Finally, as you grow further, you’ll become a director/partner responsible fortaking care of the business as a whole, leading the projects and winning clients. As for me, both choices sound equally exciting!

As you can see, depending on your preferences and seniority, there’s a wide range of projects you can get involved into! In the next post, we’ll highlight what are the core skills and mindset for a data science consultant to succeed in their role. See you on a project!

Powiązane posty

Capgemini Invent

MyWayCapgemini: Agnieszka Wierzbicka

Date icon 2022-07-21

Poznaj Capgemini Invent!

Capgemini Invent

My Way Capgemini: Szymon Kisiel

Date icon 2022-06-21

Poznaj zespół Data Science!

Capgemini Invent

What is Data Science Consulting?

Date icon 2022-04-25

Introduction to the series: “Data Science Consulting”.