All open positions

Careers · AI

Model Ops / Inference Engineer

Run model serving, release readiness, streaming, capacity and observability for EvaEsi's model stack.

About the role

What this role does.

EvaEsi is a whole new model line, built on the research from Forlais. It is a significant launch, and you can help shape it. You will work on the operational path from trained model to public product: serving, releases, capacity, monitoring, streaming and the wire between the platform and the models. This role sits beside the Lead AI Engineer and the infrastructure lead.

What you would do:

  • Run serving readiness, release operations and capacity planning for EvaEsi models.
  • Work on streaming, timeouts, cancellation, monitoring and production behaviour.
  • Help make the model stack observable and ready for launch day.
  • Coordinate with AI, platform and infrastructure on the model release path.
  • Keep the operational side of inference understandable and reliable.

What we look for:

  • Have production experience with model serving, inference systems or high-performance backend systems.
  • Like the operational side of the job: capacity, releases, monitoring and readiness.
  • Comfortable with an in-house stack built from original architecture.
  • Bring an open mind and genuine readiness for a learning curve.
  • Are results oriented and comfortable taking responsibility for outcomes.
  • Are willing to adapt and try new ideas.

Nice to have:

  • Inference optimisation, batching and caching strategies.
  • Rust, C++ or comparable systems languages.
  • Custom network protocols and streaming systems.
  • Accelerator or GPU fleet operations.

Process

How the process runs.

Each application is reviewed directly. Expect one or more conversations covering judgement, technical or craft depth, and the responsibility the role carries. EvaEsi does not publish salary ranges in public material; compensation is discussed individually during the process.