Artificial intelligence (AI) has been in the spotlight for some time now, from new, more powerful algorithms to discussions about the best infrastructure for accelerating time to results. But successfully implementing AI initiatives remains challenging. One consistently hot AI topic is how the scarcity of talent for data science-type roles hinders the number of AI projects that get implemented.
What if I could tell you that you have a hidden lever you can pull to help your AI projects succeed, and that it doesn’t require more money, hiring additional data scientists, or even buying more infrastructure?
Gap between data science and IT teams
Throughout our many conversations with customers about their advanced analytics needs, we realized that there seemed to be a gap between the data science and the IT team. Initially, this was just an anecdotal observation, so we asked market intelligence firm ESG to put their knowhow to work and investigate. They came up with fascinating results, some of which we highlight below.
When examining the challenges each team faces in working with one another, some interesting data percolated: IT’s top shortcomings were due to skill gaps, infrastructure limitations, and lack of agility. On the other hand, 40% of the organizations surveyed said the greatest challenge impacting a better partnership between data science and IT teams is the data science team’s rapidly changing priorities.
According to ESG’s research, high quality collaboration between IT and data science teams is much likelier to result in greater than expected benefits from data analytics projects.
Collaboration proves to be far more important to an organization’s ability to overachieve on their data science goals than the length of time initiatives have been underway. In fact, 88% of high-quality collaborators see increases in revenue as a direct result of data sciences initiatives versus 64% of low-quality collaborators.
Researchers also noticed that collaboration is strongest at the beginning of a project during the goals and data preparation phase. However, that diminishes during the deployment phase, which represents a substantial problem since inferencing is likely to require IT’s help with application and model deployments.
On average, organizations with high-quality collaboration achieved greater business benefits related to AI initiatives faster than those with low-quality collaboration. 87% of those high-quality collaborators have improved their business decision making process due to data science initiatives, and 88% show improvements to their customer experience.
Data scientists are an expensive and scarce resource, so you need to make sure you’re giving them the right tools to be successful. Thus, it’s important to develop a culture that maximizes the collaboration between the data science team and the IT team by putting collaboration processes in place.
For more information on the challenges that affect how data science and IT teams work, read the full ESG report.
Read Han Yang’s blog on How to Help Your Data Science Team with AI and ML
Click the following for more information on Cisco AI/ML infrastructure solutions.
Connect on Twitter: @FrancoiseBRees