Join Impress – Europe’s Leading Health-Tech Innovator

We believe everyone deserves a smile they'll love. As the largest Ortho clinic chain in Europe, we combine cutting-edge tech with expert care, revolutionizing invisible orthodontics since 2019.

With 150+ clinics across 10 countries and rapid growth, we're improving lives worldwide.

We are looking for an ML Platform Lead

to join our Onsite Team in Barcelona.

Why we're cool:

Work with an international and multicultural team

Competitive salary

Teeth aligner and whitening benefits

Collaborative work environment and positive culture

Opportunities to grow within a fast-paced, innovative company and real start-up experience with big challenges

Fresh fruits and healthy snacks at the office

What You'll Do:

Own the ML Platform: Design, build and maintain ML serving and deployment infrastructure across AWS and GCP.

Manage Cloud Infrastructure: Multi-environment setup (dev, prelive, live) using Terraform and Terragrunt.

Run ML on Kubernetes: Deployments, health checks, secrets management and observability.





Optimize GPU Inference: Profiling, batching, model conversion (ONNX/TorchScript), Triton Server.

Drive Cost Efficiency: Right-size compute, build scale-to-zero GPU strategies, own monthly infra spend and cost-per-inference metrics.

Lead the Team: Define ownership areas, grow engineers technically, set platform direction.

Bridge ML & Product: Work with researchers, PMs and clinical teams to take models from prototype to production.

What We Are Looking For:

6+ years of experience in ML engineering, MLOps, or ML platform roles with production responsibility;

Strong hands-on experience with cloud infrastructure — AWS (Lambda, Batch, Step Functions, S3, EC2, ECS/EKS, DynamoDB, RDS) and/or GCP (Cloud Run, Artifact Registry);

Proficiency with Infrastructure as Code — Terraform and/or Terragrunt for managing multi-environment cloud deployments;

Experience deploying and operating ML inference services in pro

📌 ML Platform Lead (Madrid)
🏢 Impress
📍 Madrid