From chaos to control: Building ML platform | Volvo Cars

Description

Scaling machine learning takes more than models—it takes collaboration, consistency, and control. At Volvo Cars, fragmented data and siloed teams once slowed progress. Today, a unified #MachineLearning platform is changing how teams access data, build models, and create real-world impact. In this session, George Markhulia, Engineering Manager for ML Platform, and Steve Larkin, ML Platform Engineer at Volvo Cars, share how they built an enterprise-grade ML platform that breaks down silos and speeds up innovation. They’ll walk through key architecture decisions, lessons learned, and best practices for making data easier to access, streamlining workflows, and helping teams move faster—together. You’ll leave with a practical blueprint for scaling #MLOps in complex environments, with a focus on openness, reliability, and teamwork. Learn more: If you’d like to learn more about us, click here: https://efcd.co/3QyEovk ---- Timestamps: 00:00 Intro 00:04 Introduction 01:13 Platform technology stack 02:24 Platform in numbers 03:37 Common challenges 04:57 Platform overview 06:31 Easy and difficult parts 07:35 Onboarding 11:17 Learnings 13:51 Day-to-day operations 15:27 Insights tier 16:26 ML product 20:54 Learnings 23:14 Production 29:54 Summary 32:10 Outro