Back

Speaker "Josh Bottum" Details Back

 

Topic

  • Solo Talk - Maximizing the productivity of your ML Platform with Kubeflow 1.4
  • Model Risk Management Workshop 

 

Abstract

Solo Talk - Maximizing the productivity of your ML Platform with Kubeflow 1.4

Kubeflow is an established, open source software project, which market leaders use to create enterprise-grade ML platforms with best-in-class KPIs.   Based on Kubernetes, Kubeflow enables users to streamline ML workflows and to simplify their operations.   In this talk and live software demonstration, Josh Bottum will present ML use cases and workflows that materially improve the productivity of your data scientists and ML engineers.   The talk will include an analysis of Kubeflow 1.4’s features and benefits.  The session will also include a demonstration of ML workflows, including how to build a model in an ML pipeline, and then how to efficiently tune, serve and monitor that model, which is scaled on Kubernetes pods.

MRM Workshop - 4hrs - Oct 13th 

Abstract
 
In most financial institutions, the data science and model risk/validation functions are under constant pressure to produce and update accurate governance reports. With increasing adoption of machine learning and regulatory scrutiny, the complexity of risk management and governance reporting is rapidly increasing.   The three lines of defense need to satisfy internal and external validation and audit requirements, including those defined in the Federal Reserve’s SR11-7 guidelines.    In response, modern ML Platform teams are developing automated governance workflows integrated with MLOps, which are specifically designed to improve model quality and time-to-market.  This workshop provides valuable updates, discussions and demonstrations of model risk management requirements, challenges and solutions by industry experts.   
 
 
Agenda - 4 hours, virtual 
Session 1 : Machine Learning in Banking: Applications, Model Risk, and Explainability!
Agus Sudjianto, EVP and Head of Corporate Model Risk, Wells Fargo
 
Session 2 : Risk management for pricing, forecasting and time-series models. 
Nikola Gradojevic, Professor of Finance, at the University of Guelph 
 
Session 3 : Governance-Driven Strategies for Responsible AI/ML
David Van Bruwaene, CEO-Fairly.ai
 
Session 4 : Governance Workflow Demonstration
Stuart Maiden, Data Scientist, Fairly.ai  
 
Session 5: ML Platform and governance reporting advantages of MLOps with Kubeflow
Josh Bottum, Kubeflow Community Product Manager and VP-Developer Relations, Arrikto, Inc.
 
Session 6: Model Risk Management Round Table
Agus Sudjianto, EVP, Head of Corporate Model Risk at Wells Fargo
Nikola Gradojevic, Professor of Finance, at the University of Guelph
David Van Bruwaene, CEO-Fairly.ai
Josh Bottum, Kubeflow Community Product Manager

Profile

Josh is currently a Kubeflow Community Product Manager and a VP of Community Relations at Arrikto.  Since the Kubeflow v0.5 release (in 2019), Josh has helped to improve the Kubeflow user experience by triaging GitHub issues, running user surveys, gathering requirements, updating roadmaps, writing blog posts, and conducting demonstrations and presentations. Arrikto is a major code contributor to the open-source Kubeflow Community project.