Elisabeth Günther
The Road to MLOps: How Verivox Transitioned to AWS
#1about 3 minutes
Understanding the role and challenges of MLOps
MLOps provides a structured process to build and integrate machine learning products by addressing challenges beyond just the ML code, such as versioning, security, and deployment.
#2about 4 minutes
Navigating the four phases of MLOps maturity
The MLOps maturity model guides teams through four phases: accelerating proof of concept, making processes repeatable, ensuring reliability through monitoring, and achieving scalability.
#3about 3 minutes
Overcoming siloed code and deployment bottlenecks
Verivox's initial setup suffered from siloed codebases, a lack of deployment ownership, and friction between teams, prompting a complete operational transformation.
#4about 2 minutes
Executing a multi-stage initial migration to AWS
The team's first project involved migrating from R to Python and moving from manual UI clicks to a fully automated CI/CD pipeline with infrastructure as code.
#5about 3 minutes
Building a real-time inference architecture on AWS
A standardized blueprint using Amazon SageMaker Pipelines and AWS Lambda was created to solve the major pain point of deploying models for real-time inference.
#6about 2 minutes
Using AWS Fargate for flexible batch processing
A container-based architecture with AWS Fargate and Step Functions provides the flexibility needed for custom batch jobs and lifting-and-shifting legacy projects.
#7about 4 minutes
Automating infrastructure with AWS CDK templates
AWS Cloud Development Kit (CDK) enables the creation of reusable, parameterizable infrastructure templates to scale deployments across multiple projects, accounts, and sandboxes.
#8about 3 minutes
Key learnings and results from the MLOps transformation
The migration resulted in drastically reduced deployment times, improved reliability, and new capabilities, underscoring the value of support networks and managed services.
Related jobs
Jobs that call for the skills explored in this talk.
Matching moments
42:31 MIN
Creating a stepwise transition strategy to MLOps
Effective Machine Learning - Managing Complexity with MLOps
01:58 MIN
The convergence of ML and DevOps in MLOps
AI Model Management Life Circles: ML Ops For Generative AI Models From Research to Deployment
33:17 MIN
Choosing between a custom vs managed MLOps platform
Effective Machine Learning - Managing Complexity with MLOps
26:29 MIN
Q&A on migration strategy and stakeholder management
AWS Migration within 3 months
55:32 MIN
Q&A: MLOps tools for building CI/CD pipelines
Data Science in Retail
14:42 MIN
Implementing the end-to-end MLOps pipeline
How We Built a Machine Learning-Based Recommendation System (And Survived to Tell the Tale)
18:06 MIN
Using data management and open source tools for MLOps
MLOps - What’s the deal behind it?
22:33 MIN
Automating the data pipeline with multi-cloud services
Leverage Cloud Computing Benefits with Serverless Multi-Cloud ML
Featured Partners
Related Videos
DevOps for Machine Learning
Hauke Brammer
Effective Machine Learning - Managing Complexity with MLOps
Simon Stiebellehner
DevOps for AI: running LLMs in production with Kubernetes and KubeFlow
Aarno Aukia
MLOps - What’s the deal behind it?
Nico Axtmann
Building the platform for providing ML predictions based on real-time player activity
Artem Volk & Fabian Zillgens
How E.On productionizes its AI model & Implementation of Secure Generative AI.
Kapil Gupta
Empowering Retail Through Applied Machine Learning
Christoph Fassbach & Daniel Rohr
The state of MLOps - machine learning in production at enterprise scale
Bas Geerdink
Related Articles
View all articles.gif?w=240&auto=compress,format)
.gif?w=240&auto=compress,format)
.gif?w=240&auto=compress,format)

From learning to earning
Jobs that call for the skills explored in this talk.


Senior DevOps Engineer - Search & Services - (f/m/x)
AUTO1 Group SE
Berlin, Germany
Intermediate
Senior
ELK
Terraform
Elasticsearch







Data Engineer - AWS Cloud & Data Pipelines 98% remote ID2398S
mund consulting AG
Intermediate
Gitlab
Confluence
Machine Learning