March 16, 2018 | Written by: Seth Dobrin, PhD
Categorized: Data Science | Think 2018
Share this post:
By now, in 2018, most enterprises have established some type of program to utilize math to help them make money. For many this means taking advantage of data science, or more specifically, machine learning. But despite seemingly widespread adoption, some studies show that more than two-thirds of these companies are failing to realize value from their data work.
This is largely due to a misperception that data science is an activity, or a research project. But conducting data science in a large company is fundamentally different than any Kaggle-like competition or academic project.
What organizations need is to better understand these distinctions, and to learn how to adopt and exploit them. That’s why IBM has pulled together a team of top data scientists who have personally executed data science in large companies at scale. This group, The Data Science Elite team, is a no-charge consultancy that works side-by-side with clients around a specific use-case, from inception to delivery of a model in the form of an API.
When engaging with clients the Data Science Elite team starts by working to identify a single use case. Once the use case is identified, it works with the data science leader and team to break this use case down into discrete deliverables. Most use cases worth tackling end up being tens or hundreds of individual models that are either run separately or together as an ensemble.
The team’s delivery mechanism is based on agile principles. This means that each model will be delivered as a series of sprints. The team will engage with client data scientists to break down the delivery of the model into 3-4 sprits with the aim of beginning to deliver value in as early in the process as possible.
Since we are using this team to help clients learn our process, we require clients to at least match the resources that IBM’s Data Science Elite team commits. This is in the form of a subject matter expert, data scientists and data engineers.
Based on years of hands-on work with clients, we know that a broad spectrum of maturities and understanding around data science exists. As a result each client has a different entry point for advancing their efforts. To help position clients, we’ve categorized three distinct categories in which most organizations involved with data science exist, Mature, Nascent, and Aspiring, each receiving a more intense level of support.
The commitment IBM is making in helping our clients is significant. Looking at the different levels of maturity IBM’s commitment at typical rates for this skill set vary based on the maturity of the clients data science program, but can range up to $200,000.
IBM is committed to helping our clients and potential clients be successful implanting data science into their organization primed to be an integral part of how they do business so data science is not a scam for them. To learn more about the data science elite team, download this one page overview and reach out to your IBM sales team.
- Photo: Six members of the newly minted Data Science Elite Team: (L-R) Annamaria Balazs, Umit Cakmak, Seth Dobrin, Susara van den Heever, Wendy Wang, and Siva Anne (Credit: Mike Webb Photography)
THINK Blog: Empowering the New Data Developer