As an MLOps Engineer at United Algorithmics, you will bridge the gap between research and production, ensuring that machine learning models are reliably trained, deployed, and monitored.
Responsibilities
- Build and maintain training infrastructure for large-scale ML workloads
- Design model serving pipelines with low latency and high availability
- Implement continuous evaluation frameworks and model performance monitoring
- Automate the ML lifecycle from experimentation to production
- Collaborate with research scientists to operationalise new models
Requirements
- Experience deploying ML models to production (BentoML, Triton, or Ray Serve)
- Proficiency with ML frameworks (PyTorch, JAX) and experiment tracking (MLflow, W&B)
- Strong Python and infrastructure engineering skills
- Experience with feature stores, data pipelines, and model registries
- Knowledge of LLM serving and inference optimisation is a plus
