Deploying a Serverless Model on AWS SageMaker – A Complete Walkthrough
Complete walkthrough of training and deploying a serverless machine learning model on AWS SageMaker with cost-effective infrastructure
Read more →About Me
Hi, I'm Anton Roesler — an MLOps Engineer and Software Developer with a passion for building robust and scalable machine learning systems that work in the real world.
In my role at Bosch, I've worked across a wide range of machine learning use cases — helping product teams and data scientists move from experimentation to production-ready ML applications. My focus is on engineering solutions that don't just run in notebooks — but operate reliably at scale, in the cloud, under real-world conditions.
Areas of Expertise
-
ML CI/CD
Automated, testable pipelines for training, evaluation, approval workflows and deployment
-
Model Monitoring
Metrics, alerts, and drift detection for live models in production
-
ML Infrastructure
Design and implementation of scalable AWS- and Azure-based training and inference systems
-
Containerization
Docker-based development and reproducibility workflows, optimized for team collaboration
-
Modular ML Systems
Architecture of maintainable systems: decoupled data ingestion, preprocessing, and inference modules
-
Cloud-Native Deployment
Use of Terraform, GitHub Actions, SageMaker Unified Studio, EKS or Lambda for scalable operations