Accelerate business value with a scalable data science platform

By Carlo Appugliese
@CAppugliese

While data science and AI are relatively new in the market, the concept of extracting value from data has been around for a while. But time has changed things —  and leaders are dealing with a variety of tools (open source options with Rand Python and proprietary packages, like SAS and SPSS), skills and techniques (machine learning, optimization, stats and more) and a variety of workload options dealing with large data (EDW, Hadoop and more). Leading  data science teams is no easy feat — and many have been asking questions about how they can maximize their business value today. Carlo Appugliese shares key insights from his team's experiences helping hundreds of companies enhance their data science practice, ranging from how to shorten time to impact to how to extract the best work from your team.

Carlo shares six steps to leading an effective DS team:

1. First and most important, focus on your business objectives and problem; don’t get caught up with technology. Don’t do technology for the sake of technology.
2. Don’t overanalyze what you’re trying to do. Some of the most successful projects are delivered by just doing it. Identify your business problem/use case, put a working team together and go.
3. Don’t limit the data your team has available to them. Data scientists need to explore, and sometimes their approach will change based on findings. Give your scientist access to all the data.
4. Think about how you will operationalize your project at the start. Don’t code an entire solution and then rewrite the model in a digital application. Build the solution in a technology you can operationalize.
5. Don’t be afraid to use the latest Python frameworks. Open frameworks are community-developed and have outperformed traditional approaches. Just scan some the Kaggle winners, http://blog.kaggle.com/category/winners-interviews/. Python is the clear leader, so don’t be afraid of open source.
6. Empower your team, so that even the most junior scientists can use open frameworks to solve your business problems. Some of my most productive team members are knocking it out of the park right out of graduate programs.