Last week, the Seldon team headed to the Beurs van Berlage in Amsterdam for the return of the World Summit AI. Excitingly, this was our first in-person conference since the start of the pandemic! World Summit AI is internationally renowned as the world’s leading conference on artificial intelligence, bringing together the global AI ecosystem to discuss the technology’s most pressing issues and to set the global agenda around its use. It was a pleasure to take part, hear from some of the brightest brains in the industry, meet some of our European customers and join the UK and Dutch Embassies for a lunch to discuss AI strategies and joint initiatives.
Seldon’s Lead Solutions Engineer, Tom Farrand, joined the lineup of speakers for a hands-on workshop on deploying and managing machine learning (ML) models at scale. Tom leads our solutions engineers at Seldon, where he knits together client needs with Seldon’s deployment capabilities to help our customers realise the potential of ML.
In his talk, Tom discussed what organisations need to understand and have in place, in order to deploy ML models at scale while minimising risk. Tom took the audience through the creation, deployment, monitoring and management of an ML model using a real-life use case.
He began his talk by highlighting the key challenges that organisations face when scaling up their ML deployment, why businesses should care about MLOps and how to overcome them using Seldon’s technology.
Tom spoke on the various stages of deployment, and the various technical challenges that can arise when managing more than a handful of machine learning models. Without sufficient monitoring or insight, these challenges can impact ROI, cause slow workflows between data science and DevOps teams, and risk reputational damage and financial loss.
Tom then highlighted the importance for organisations to leverage MLOps tools to solve various challenges, to increase cost savings of 64% on average for each subsequent year with productivity gains – meaning that organisations can deploy their machine learning models in days, rather than months. Furthermore, Tom introduced Seldon Deploy and the four key aspects of its functionalities: to serve, monitor, manage and explain models and workflows at scale and quickly.
The talk was followed by a hands-on workshop, where Tom covered a range of practical tasks including training models on pre-processed datasets, deploying model artefacts with Seldon Deploy, and using explainers and outlier detection to gain insight into model decisioning.
During the event, our VP of Alliances Keith Bush also attended a lunch with the UK and Dutch Embassies to discuss AI strategies, identifying building blocks and barriers to AI dissemination across joint initiatives such as skills, talent and R&D. Attendees included members from the UK and Holland, academia, private sector companies such as Shell and Unilever, as well as UK-based machine learning startups such as Mind Foundry and Seldon.
We were also joined at the conference by some of our European customers: Philips, Flo Health and Ahold Delhaize, and discussed how they’re working with our Seldon product suite to deploy their ML models faster and more efficiently. Meanwhile, we networked with the global AI community and listened to talks from some of the biggest names in technology on the work they’re doing with AI and how they’re using it as a force for good. It was great to be back, and we can’t wait to do more in-person conferences in the coming months.