The chief data scientist: Who they are and what they do

• Read original article here

Anodot’s Ira Cohen shares what it means to be a chief data scientist, particularly in the era of Covid-19.

We’ve been exploring working in data for some time now at Silicon Republic. We’ve looked at the difference between a data scientist and a data analyst, the types of skills data professionals need and how much money you can expect to make in the industry. But something we haven’t discussed in depth is the role of a chief data scientist.

Ira Cohen is a co-founder and the chief data scientist at Anodot, a US analytics company that helps businesses detect anomalies in revenue, customer interactions and more. He draws on AI and machine learning to invent and develop real-time algorithms to carry out this detection.

Before co-founding Anodot, Cohen worked as the chief data scientist at Hewlett-Packard Enterprise. Here, he explains the differences between his role and that of a CTO, and why chief data scientists will be critical in navigating the aftermath of Covid-19.

The chief data scientist manages a range of data-driven functions including overseeing data management, creating data strategy and improving data quality. They also help their organisations extract the most valuable and relevant insights from their data, leveraging data analytics and business intelligence.

Perhaps most importantly, organisations rely on the chief data scientist to bridge the gap between management and the data science teams, helping them understand what machine learning can achieve and, conversely, not accomplish. The chief data scientist has a much deeper understanding of these technologies than the CTO, who likely has a broader knowledge base but not necessarily the deep expertise.

Machine learning is a remarkable innovation when supported by large amounts of data but the journey from big data ideas to successful machine learning implementations is often a complex and arduous one. This path requires a trusted navigator who can help the data science team overcome potential challenges – that is where the chief data scientist comes in.

Experienced chief data scientists understand that data is the fuel behind key initiatives and know the non-deterministic risk of developing these capabilities. They bridge the gap between organisational expectations and the reality of what machine learning can accomplish, while understanding how to mitigate the risks associated with complex data-driven endeavours.

Many organisations are discovering that they really need a chief data scientist. For data-driven organisations, this role has become a must-have position rather than a luxury.

Since Covid-19 we have seen the rise of the chief data scientist, especially as organisations accelerate their digital transformations. Right now, everyone is engaging customers and partners in different ways in the digital world, launching new business models and finding better ways to bring products and services to market. This has led organisations to embrace more ambitious data strategies that require more experienced data science leadership.

We are seeing chief data scientists become heavily involved with board-level and c-suite-driven corporate strategies as data becomes even more central to critical company decisions. IDC recently completed a survey that revealed that 59pc of chief data scientists now report to their CEO or another c-suite executive, which illustrates just how far this role has come in a short period of time.

One of the most important things a chief data scientist can do over the next few months is to use machine learning to solve the most pressing business problems created by Covd-19 and the global recession.

For example, churn prediction is a key dilemma for organisations right now – specifically, they must forecast which customers are most likely to stop being customers. This important task requires superior data analysis know-how. Moreover, assessing churn predictions requires different levels of technical and data science expertise – qualities that the chief data scientist already has.

For example, some organisations need to predict churn in real time, while others must assess churn for each customer once a month. This requires different expertise and product sets with unique machine learning requirements.

Having a chief data scientist navigate these varied scenarios would likely deliver positive outcomes as that individual could understand the scope of the data science work required and complete a thorough analysis of the approaches that will or won’t work – all while balancing the business and technical trade-offs of taking one approach over the other.

This level of understanding will make all the difference when it comes to finding the most effective solutions that yield the desired results under the right cost and time parameters.

Truly great chief data scientists know how to walk a fine line between driving creative innovation and pragmatic solutions. As data scientists are researchers at heart, they need the time and space to explore different problem sets and possible data-driven solutions. At the same time, they must also deliver real-world data management solutions that solve their organisation’s pressing business problems. The ideal chief data scientist knows how to rally teams to deliver both.

For researchers, it is all too easy to go down rabbit holes searching for the best solutions. Sometimes you find the gold, but many times you do not uncover it. Talented and resourceful chief data scientists know when to pull their teams out of the rabbit holes; when they have asked all the right questions and done the hard work but still cannot find the treasure. That is when they must be pulled out to avoid wasting too much time, and then you can move them on to the next hole.

Many organisations let their data science teams spend too much time with their heads buried in rabbit holes that end up not bearing any fruit. Finding the right balance between exploration and pragmatic solutions is a key role for the chief data scientist.

Machine learning is the most important technology for chief data science officers in the year ahead.

Machine learning is particularly critical for data science professionals right now. Many of them are engaged in a build-versus-buy debate regarding products or services that offer machine learning as a core feature.

Organisations that opt to build their own platforms with machine learning capabilities should only do so if they are creating mission-critical applications. Otherwise, they will likely spend a great deal of time and expense building an internal technology that will not deliver as much value as the time and effort they put into it.