David is a Lead Data Scientist consultant at Computas who specialises in machine learning and MLOps in particular, helping companies scale their ML initiatives and taking them to production. He has worked on such projects for a range of customers in different sectors, recently for Ruter. He has a background in computational physics and has previously studied friction and climates of other planets as a researcher in academia.
Ruter plans, coordinates, orders and markets public transport in Oslo and former Akershus (now part of Viken county), and is actively using machine learning to improve its services. This includes forecasting passenger numbers for over half a million departures every day, which are shown to customers in the Ruter App.
The first model was deployed in the summer of 2020, but updating it was time consuming and monitoring was limited and manual. In this talk I will present how Ruter is actively pursuing MLOps best practices to support machine learning models in production, making sure models are reproducible, easily updatable and extensively monitored.
Q & A with David Skålid Amundsen
1. What is your full name and title?
David Skålid Amundsen, PhD
2. What is your occupation?
Lead Data Scientist at Computas
3. Tell us about your first encounter with artificial intelligence (AI)?
My first memorable encounter with AI was probably when I heard on the news that IBM’s Deep Blue had beaten Kasparov in chess.
4. What is your competence within the field of AI?
I am a Data Scientist, sometimes developing machine learning models, but often helping companies streamline and robustify their ML efforts.
5. Why did you develop an interest in AI?
Having a background in computational physics, I was fascinated by the increasing application of algorithms that learn directly from data.
6. Can you recommend a relevant book or film about AI?
I enjoyed watching AlphaGo, depicting the match between the top Go player Lee Sedol and AlphaGo developed by DeepMind.
7. Why should we, or should we not, be afraid of AI?
AI has enormous potential, but we should certainly be careful and cautious with respect to fairness, transparency and its impact on society.
8. Which field, in your opinion, has the most to benefit from AI – and why?
That would depend on how you measure benefit, but applied AI in the medical field has the potential to directly improve people’s lives for example in terms of better diagnostics and medicines.
9. How should the use of AI develop in the future?
Use of AI should develop with a focus on transparency and fairness, with regulations governing how it can be used particularly when it directly impacts people’s lives.
10. Why should participants tune in on your presentation during AI+?
To learn about how Ruter has gone from an AI proof of concept to a production system serving over half a million passenger volume predictions to hundreds of thousands of travellers every day.