In my role as VP of Enterprise AI & Data Science Solutions at Future Tech Enterprise, Inc., I hear a common question from our customers: “Who should own AI and data science in my organization?” Should it be the business’ leadership? Should it be IT? Or should it run as its own domain of data scientists?
My answer is simple: whatever works best for your organization.
It is less important who “owns” AI and data science than it is who is “part of it.” A successful AI and data science practice is a coordinated approach that has professionals bringing their specific areas of expertise and concerns together to use the latest tools to solve critical business challenges. Data science “ownership” is less important than ensuring active, engaged participation across all of the organization’s stakeholders including business, IT, security, strategy, compliance.
Collaboration across all of these functions is essential. It will help ensure your company has a successful, enterprise AI and data science program that delivers valuable insights, improves internal processes and customer experiences, and generates substantial business benefits. The “ownership” of data science should be the result of a collaboration between business and technology, ideally within an enterprise data management framework.
If clarity, focus, and collaboration are organic within your organization, ownership is less critical than participation. If not, consider mapping data science to a top executive who can champion AI, machine learning (ML), and business analytics across the entire organization to stimulate that sense of accountability.
Having the data science team “start at the top” is helpful in showing executive sponsorship backing, which is often needed to swiftly clear conflicts (e.g. priority, resource, turf, etc.) and ensure that everyone stays problem- and business-focused.
How can you tell if you have a good fit? Watch the team, their attitudes as much as their performance. Most data scientists are much like engineers at heart: they live for the challenge of not only solving the problem but doing so elegantly. So, if they are focused on the issue and talking (even arguing!) about the best ways to solve it, things are good. Better still, when they are engaging with the larger team and presentations are more unified, seamless, and feel less like a compilation of individual sections, you have a well-integrated, high functioning team.
It is when they are talking about schedules, meetings, processes, and procedures that might warrant a closer look. True – these things are important as well – but they are typically not something that data scientists deal with directly, especially if there are other team members (e.g., mission assurance, QA, business process / change management groups, etc.) that can solve this.
Keeping data scientists focused on doing actual data science is the best indicator that your team is working optimally. It’s also a great way to retain that talent – happy employees tend not to look for other opportunities. The data science war for talent is very much a real thing right now and it’s impacting nearly every industry. How will your organization tackle this war for talent?
You can learn more about how Future Tech and Dell Technologies are working together to optimize AI and data science processes at the federal level here.
