When setting up your data science projects and organization, the goal is to make the right initial investments and establish momentum quickly. Unfortunately, corporations can falter at this stage by either being too rigid and creating a strategy before they really understand data science or being too flexible and setting up multiple teams and proofs of concept without clear objectives and alignment. This talk and presentation will give leaders and contributors alike a set of clear strategies and focus areas for building their data science teams successfully:
1. Create the right team: The team you bring together will either make or break your data science endeavors. While it is easiest for organizations to often bring together a combination of available resources, they also need to evaluate if those resources are best-suited for the challenge. Each member of the team needs to be hand-picked and should understand the overall objective and the role they need to play.
2. Select the right use case: Use case discussions often start around the art of the possible but eventually need to evolve into the art of the valuable.There are three questions that a strong and early use case can successfully answer:
1) Is there measurable value that can be demonstrated to the business?
2) Is the use case a feasible and quick win?
3) Does your team have the right resources to succeed?
3. Innovation: Because of the rapidly evolving nature of data science, a culture of innovation is critical to sustain success in data science. Data scientists often get bogged down and spend an inordinate amount of time with inefficient, distracting, and frustrating business processes, technologies, or data. In order to deliver and sustain business value while also retaining talent, best practices and strategies on fostering innovation will be shared.
Head of Data Engineering and Analytics, Chewy