As a data scientist, it is paramount to establish cross-functional teams who are champions of what you are building. In this talk, we address common challenges and what to do about them, with a focus on data science methodologies, patterns, and best practices. We also address how to structure and position your team to successfully move a project from an exploratory phase to production-facing.
Even after successful exploratory analysis, a vast majority of data science projects still do not make it to production. No matter how outstanding the technical merit and capability, data science projects must also focus and deliver in two key areas – people and process – to truly make a valuable business impact. It takes an entire team who is educated, accountable, and focused on driving value to deliver a data science project past an exploratory phase and into production.
In this talk, you will learn about the challenges of building data science solutions in large enterprises, including the criticality and methods to achieve buy-in across all levels of an organization. Then you will learn about the key differences between exploratory and production model building approaches, which focus on engaging the right team and adapting techniques for engagement. We will wrap up by walking through a real-world example, showcasing a recent project focused on anomaly detection with IoT data.