PyData Montreal Meetup #21 (in-person | en personne)

PyData Montreal Meetup #21 (in-person | en personne)

COVID rules | Règles COVID
EN: Wearing masks is not obligatory but encouraged when not consuming drinks or food
FR: Le port du masque n'est pas obligatoire mais encouragé lorsque vous ne consommez pas de boissons ou de nourriture

Agenda

• 5:30 pm - Doors open
• 5:50 pm - Introduction
• 6:00 pm - "Introduction to Trustworthy Machine Learning for the Enterprise" by Mohamed Leila, ServiceNow
• 6:40 pm - Break and networking (coffee, tea, soft drinks, and pastries will be provided)
• 6:50 pm - "ML in production in the video game industry: Ubisoft's use case" by Jean-Michel Daignan, Ubisoft
• 7:30 pm - Final notes, break, and networking
• 7:45 pm - Drinks & social hour at Bar Pamplemousse, 1579 St Laurent Blvd

Ordre du jour

• 17h30 - Ouverture des portes
• 17h50 - Présentation
• 18h00 - "Introduction to Trustworthy Machine Learning for the Enterprise" par Mohamed Leila, ServiceNow
• 18h40 - Pause et réseautage (café, thé, boissons gazeuses et viennoiseries seront offerts)
• 18h50 -"ML in production in the video game industry: Ubisoft's use case" par Jean-Michel Daignan, Ubisoft
• 19h30 - Notes finales, pause et réseautage
• 19h45 - Boissons et heure sociale dans Bar Pamplemousse, 1579 St Laurent Blvd

***

Introduction to Trustworthy Machine Learning for the Enterprise
Abstract:
The term "Trustworthy Machine Learning" refers to a collection of frameworks and techniques designed to understand and mitigate the unforeseen risks associated with Machine Learning Systems. With the proliferation of use cases for Data Science and Machine Learning in the modern Enterprise, there is a growing need to integrate Trustworthiness and Governance in the Machine Learning Development Process. This talk presents an overview of Trustworthy Machine Learning and discusses some best practices for Enterprises to ensure Trust & Governance is an integral part of their AI Journey.

About Mohamed:
Currently, he's the Lead of the AI Trust & Governance Lab at ServiceNow. Their vision is to establish ServiceNow as a leader in Trustworthy Enterprise AI, known for its Robustness, Transparency, Privacy, and Fairness. Prior to this role, he led an effort to investigate the feasibility of Privacy-Preserving Machine Learning technologies such as Federated Learning for some use cases at ServiceNow.
Before joining ServiceNow, he worked in multiple roles in Data Science, Engineering, and Operations Research. Mohamed received his Ph.D. from McGill University in 2017, where he conducted research on Optimizing Biofuel Supply Chains.

ML in production in the video game industry: Ubisoft's use case
Abstract:
Like other companies in the video game industry, Ubisoft is facing the challenges of deploying and operating machine learning systems that should scale and support different use cases. This presentation will be the occasion to present the reality to deploy machine learning in production for Ubisoft by presenting his internal ML platform Merlin with also some feedback on various tools of the MLOps world that are used or tried.

About Jean-Michel:
He has been a data scientist in the Ubisoft data department for four years; he is operating recommender systems in-game, helping the development of the Ubisoft internal ML platform as a glorified guinea pig, and supporting Merlin's user. He previously worked in various companies like EDF and CEA as an R&D engineer/data scientist.

***