Identifying and Prototyping Data Science Use Cases
While nobody questions the value of “AI” or data science these days, most organisations, leaders but sometimes even data scientist cannot tell where and how to focus its power to put it to the best use. There is simply not enough experience in that around yet. This is the first topic we’ll be focusing on: how to identify and prioritise possible use cases for a data science team using a structured methodology based on what’s feasible and where the highest value can be realized.
Secondly we’ll look into how can we build an MVP or prototype in the shortest time to prove an idea’s value. We’ll cover specifics like data and infrastructure requirements for certain tasks types, team composition, when to use Scrum and when to not and why does it make sense to think in a platform and keep the route to deployment in mind early on.
On the practical side we’ll highlight how self-service, visual data pipeline tools can enable the citizen data scientist to build such prototypes on its own.
Senior Manager, Enterprise Analytics Consulting, EPAM Systems
Ambrus is a passionate analytics consultant with over a decade of international experience covering a wide range of industries and solutions, having worked with both mid-sized and leading global organisations. He always implemented solutions end-to-end as in his mind the domains of business analysis, solution and interface design and organisational transformation are all connected.