Earlier this month we held the first edition of AI in Practice at the Startup Village in Amsterdam Science Park, in collaboration with Capgemini Invent and IBM. We wanted to create an event that breaks through the Artificial Intelligence hype and showcases the real business value that’s already emerging from AI applications across sectors. We heard from a diverse range of speakers – from law firms and logistics companies through to medical imaging and media companies – all sharing their experiences with AI.
Inspiring use cases
Optimising express delivery
Luc Jeurissen a Program Manager at DHL discussed a recent project within the express department of DHL. The geographical location of delivery points was frequently inaccurate and the delivery windows very broad, leading to a high number of undelivered packages. In just four months DHL improved their geocoding accuracy from 30% to 98%. As Luc explained, “using AI is about improving the user experience, not just cost saving.”
Optimising stroke care with AI
Renan Sales Barros the CTO at Nico.lab started by sharing some sombering facts about strokes. Did you know that one in six people will suffer a stroke in their lifetime? Once a stroke has taken place, the longer it takes to be diagnosed and treated the increasing chance that the patient will be left with a permanent disability. Nico.lab have developed an AI platform that analyses brain CT scans in less than 3 minutes, dramatically increasing diagnostic process and the chances of a full recovery.
Creating personalised playlists
Due to a growing customer preference for on demand media consumption, FD Mediagroep wanted to offer its audience a curated playlist of news content. This enables listeners to select preferred topics and receive a tailored supply of the latest news segments. Philippe Bressers, FD Mediagroeps AI Product Owner talked us through the development process and gave a live demo of how the app works.
Accelerating the breast cancer diagnostic process
Florin Manaila a Deep Learning Senior IT Architect at IBM shared some fascinating research he’s been undertaking on breast cancer detection with IBM PowerAI Vision. His focus was on answering the following question – can we accelerate the diagnostic process with deep learning? The results are very promising with 93% accuracy in identifying the types of tumors with deep learning.
Below are some of the commonalities that came to light as the afternoon went on.
– Take the time to define what problem you’re trying to solve with AI
– You need data, data and more data!
– Quickly release, fail and learn
– It’s not a project, it’s a process – invest in your people and team
– Partner with expert parties to kickstart the planning
We’re already looking forward to the next event on the 6th February 2019. Secure your place at www.aiinpractice.nl