The Balancing Act of AI in The Enterprise and The Rabbit Hole of a Medical-Aid Application (South Gallery, Room 6)
Information
To get to South Gallery Room 9, please follow the floorplan arrow (between Digital Transformation Acceleration and POC to Production) to the South Gallery Rooms. Then take the elevator to Floor 2 and follow the signage.
Designing and implementing AI products in an enterprise is an endeavour that requires harmonizing antagonistic dimensions, like trying to collect astonishing amounts of data while respecting privacy, automating experiences without alienating consumers when something goes wrong, or leveraging powerful compute capabilities while keeping applications accessible to populations with low power devices. When talking about these issues, we risk falling in platitudes or using too much business-speak or buzzwords, so we will anchor ourselves to a concrete example on how we dealt with these issues, and the complexities of delivering the actual AI product, for an application to assist dermatologists on assessing and treating psoriasis
3 Key Takeaways
- Balancing problem framing, application functionality, deployment and modelling decisions as a whole
- Solving technical constraints on low-data domains
- How to create data feedback loops and roadmaps for product evolution
