Responsible ML Innovation and Data Product Management (a WiDS Zurich event)

This event is part of the WiDS Zurich (www.wids.ch) conference.

Over a decade into the current "AI summer", the adoption (and, hence, much-expected impact) of ML across various industries is more relevant than ever. Let's talk about two of its most pressing aspects:
How to increase the share of promising prototypes that become productive? AND
How to enable responsible ML by design?

Agenda for the evening:

17.30 – Doors Open
18.00 – Welcome and Opening
18.05 – Driving ML Innovation with Data Product Management (Dr. Andreea Hossmann)
18.35 – Responsible ML Is Explainable ML (Natalie Bolon-Brun)
19.05 – Closing
19.10 – Apéro

Dr. Andreea Hossmann, Principal Product Manager, Data, Analytics & AI, Swisscom

Andreea Hossmann is responsible for the successful creation of data products as Principal Product Manager for Data, Analytics and AI at Swisscom. She is also a Venture Associate, working with Swisscom Ventures to assess AI startups worldwide. Prior to her Product Management career, Andreea was a Senior Data Scientist, before assembling and leading a Data Science team to work on AI topics, such as natural language understanding and search. She is an experienced researcher with a background in applied machine learning, network science and computer networking from her PhD education at ETH Zürich.

Abstract:
Machine Learning is the logical continuation of the "Software Eating the World" trend. At Swisscom and DNA, we are already touching upon data and ML use cases along the entire value chain, from marketing and sales to network and operations, including support functions. But we strongly believe we can do even more! Our ambitious vision for the future is to spread the data and ML skills, expertise, and mindset from the current 42 agile product teams of DNA to the entire more than 1000 teams of the whole company, and truly make ML accessible to everyone. In this talk, I will touch upon different strategies we're employing to reach that vision.

Natalie Bolon-Brun, Data, Analytics & AI Engineer, Swisscom

Natalie Bolon Brun is a data scientist on the Flywheel team at Data, Analytics and AI at Swisscom. Her work verses around customer understanding with special interesting on classification for unbalanced data and explainability for deep learning models. Prior to joining Swisscom, Natalie received her Masters in Electrical Engineering from EPFL.

Abstract:
As data-driven company, Swisscom is incorporating more and more machine learning models to leverage value from the large amounts of data collected every day. Nevertheless, the advance towards more complex models for better performance usually comes at the cost of lower to no interpretability. In this talk we will explore the area of explainability for machine learning with a focus on deep learning models. We will go through an overview of current methods on the field and go through an example of how Swisscom is approaching the topic.