Description :

At Upsurge Labs, we work across AI, robotics, biotech, and new computational fields. We need MLOps Engineers who can keep intelligent systems running smoothly, scale them efficiently, and help them evolve.

Responsibilities :

- Work methodically, organised, and value precision and process.

- Approach pipeline design carefully, balancing reliability and flexibility.

- Enjoy working where DevOps meets AI research, solving complex problems.

- Focus on automation, optimisation, and reproducibility in workflows.

- Be persistent and disciplined, able to troubleshoot issues even during off-hours and document clearly.

- Understand this is a fast-moving and complex field, but that's where top engineers excel.

Requirements :

- Strong understanding of end-to-end ML lifecycle - data ingestion, training, deployment, monitoring, and continuous retraining.

- Expertise in MLOps tools and frameworks - MLflow, Kubeflow, Airflow, Metaflow, Vertex AI, SageMaker, and Weights & Biases.

- Proficiency in Python and cloud infrastructure (AWS, GCP, Azure) for AI workloads.

- Experience with Docker, Kubernetes, and Helm for scalable model deployment.

- Familiarity with feature stores (Feast, Tecton, Databricks Feature Store) and data versioning tools (DVC, LakeFS, Delta Lake).

- Deep knowledge of CI/CD for ML systems and infrastructure as code (Terraform, Pulumi).

- Understanding of model serving architectures - TensorRT, TorchServe, vLLM, FastAPI, BentoML, Ray Serve.

- Proficiency in observability and monitoring (Prometheus, Grafana, OpenTelemetry, Seldon Core).

- Experience with vector databases, embedding stores, and LLM pipeline orchestration (LangChain, LlamaIndex, Prefect).

- Awareness of AI model governance, security, and ethical compliance frameworks.

- Bonus : Familiarity with GPU cluster management, distributed training (DeepSpeed, Horovod, PyTorch FSDP), or on-device ML (Edge Impulse, TensorFlow Lite Micro).