ML Engineer - MLOps & Platform Engineering

You build and own the platform behind our AI Solver: the systems that manage our training data and models as first-class assets, bring them reliably into production, and serve them to customers.

• Build the training and data platform. Design the pipelines and systems that version, track, and manage our training data and models as the assets they are, with reproducibility and lineage built in.
• Own the model lifecycle. Build the path from experiment to production: model versioning, a registry, promotion, and reliable, repeatable training and deployment.
• Close the loop tp production. Build the monitoring that surfaces model degradation and flags when incoming data drifts outside what a model handles well, so our AI engineers know where to act.
• Enable the AI team. Provide the workflows and tooing our AI engineers and data scientists use to train, evaluate, and deploy models. You build the rails, they drive.
• Rund and evolve the production service. Operate and scale our AWS service that serves the AI models, keep it fast and reliable, and extend it as we grow, for example from serving a single model to multiple selectable models, including access-controlled or user-specific ones.
• Work hand in hand with the Cloud team. They build our simulation platform and are the main consumer of your AI service, so shipping new capabilities means designing the interface and rollout together.
• Pitch in where it counts. We’re a smal lteam, so the platform work reaches into classic software and infrastructure engineering. You’ll have room to follow the problem whereever it leads.

This role builds and runs the platform. Assessing model quality, curating training data, and the modeling itself sit with our AI engineers and data scientists. Your job is to make their work fast, reproducible, and production-ready., * Deploying AI models beyond the cloud: CPU-only on-premises or edge targets, and hybrid setups.

• Workflow orchestration (Airflow, Prefect, or similar).
• Inference optimization (quantization, pruning, efficient architectures).
• AWS stack (S3, EC2, ECR, SageMaker) and infrastructure as code (Terraform).
• Building internal platforms or tooling that other engineers build on.

You won’t check every box. If you know your gaps and how to close them, apply.

Why us?

• A real technical challenge. You’re reshaping a proven simulation engine for a market moving to cloud and AI.
• Ownership and impact. About 40 people. Your decisions shape the product and the business.
• Modern tooling. Notion, GitHub, Linear, coding agents. We’re building the practices that make this work, and you help shape them.
• Direct and honest culture. Candid feedback is standard practice for us, both internally and externally. No micromanagement.

• Background in Computer Science, Data Engineering, Machine Learning, or a related field, with 3+ years of relevant experience. We’re hiring at mid to senior level.
• Strong Python skills and solid software engineering fundamentals (testing, version control, CI/CD).
• Hands-on experience taking ML systems from training into production: data pipelines, training workflows, and deployment.
• Experience with cloud environments and containerization (AWS, Docker, Kubernetes, or similar).
• Familiarity with experiment tracking and model/data versioning tools (e.g., MLflow, Weights & Biases, DVC).
• Pragmatic and reliability-minded. You focus on building systems that work and keep working.
• Coding agents are part of how you build, and you treat them as a system to optimize, not a gadget you occasionally reach for. You keep sharpening how you work with them, from context and tooling to workflow, and you know exactly where they help and where they get in the way.
• English is our working language and all you need to do the job. German is a plus. We’re still a mostly German-speaking culture shifting toward English.

For decades, Cadmould has been one of the fastest and most advanced injection molding simulators on the market, trusted across the plastics industry. Then we built something the field had never seen.

Cadmould AI Solver is the first Large Engineering Model (LEM) for plastic injection molding: a transformer-based neural physics model that delivers high-fidelity results up to 1,000x faster than classical solvers. It turns simulation from a slow validation step into something engineers can explore in real time. It’s live as a research preview on our site, and it has already shipped to our first customers.

Powerful models are only as good as the data they learn from, and only matter once they ship. That’s your domain. You’ll treat training data as a first-class asset: versioned, traceable, and continuously improved, with its impact on results made visible. You’ll build the pipelines, the model lifecycle, and the live AWS service that carry our models from experiment to customers. The systems around the models are as much the product as the models themselves.

This role can be performed from our office in Würselen near Aachen or remotely from Germany, with occasional travel for team events and on-sites.