In today’s world, where data and artificial intelligence are transforming industries, forming an ML platform team is crucial for organizations seeking to scale and fully leverage AI. But what makes an effective ML platform team? When talking about platforms, it might seem that the provided tech stack is the most significant area of focus for such a team. While that is important, I believe that the key factors for success actually lie elsewhere: in user experience and stakeholder engagement. This speech aims to demystify the role of an ML platform team, offering a personal viewpoint on when an organization should consider setting one up and the importance of such an investment. I will also provide practical insights into how we have established and run the daily operations of the ML platform team at Wolt.
Miia Niemelä is a Machine Learning Engineer for Wolt’s ML Platform team. In her role she helps other teams in building their ML products and develops tooling for them, recently she’s been developing tooling for feature storing. Miia believes that the most important purpose of an ML platform team is to ease the life of data scientists and speed up the delivery of good quality ML products, and therefore the user experience of the tooling needs to be always kept in mind. Previously Miia has worked as a consultant in data & ML engineering, helping customers to plan their ML platform architectures and bring DevOps mindset to the development of ML products.
