Overview
Unico Connect is seeking a Senior MLOps Engineer to architect and operate the ML platform that powers AI delivery across client engagements. The role designs deployment patterns, observability stacks, and the AI-engineering platform that takes models from proof of concept to enterprise production. The position is well suited to senior platform engineers with deep cloud experience, hands-on Kubernetes expertise, and a track record of building developer-productivity wins for ML teams.
Key Responsibilities
- Architect the ML platform: training pipelines, model registry, feature store, deployment, and monitoring
- Design LLM ops infrastructure: prompt versioning, evaluation pipelines, cost dashboards, and fallback routing
- Lead model deployment on GCP Vertex AI, AWS Bedrock and SageMaker, or self-hosted platforms (Modal, Ray Serve, vLLM)
- Mentor 1-2 MLOps Engineers; set MLOps standards and review pull requests against them
- Drive cost optimisation through model tiering, caching strategies, batch inference, and spot instance usage
- Partner with AI Engineers and the Tech Lead on production-readiness reviews
- Govern compliance posture for ML platform components including SOC 2, HIPAA, and GDPR requirements
- Lead capacity planning for GPU and inference workloads
- Design rollback, canary, and shadow-traffic patterns for risky model deployments
- Author internal MLOps playbooks covering incident response, drift remediation, and cost emergencies
- Run quarterly platform retrospectives and roadmap reviews
Required Qualifications
- 5+ years of MLOps or platform engineering with at least 1 production ML or LLM platform shipped
- Deep cloud experience on GCP, AWS, or both at production scale
- Strong Python; comfortable working through PyTorch, TensorFlow, or JAX internals
- Kubernetes production experience including operators (Kubeflow, Argo)
- Hands-on with model serving (Triton, vLLM, TGI, KServe, or Ray Serve)
- Strong fundamentals in observability (Prometheus, Grafana, OpenTelemetry)
- Track record of mentoring junior MLOps and DevOps engineers
- Capacity to lead cost-optimisation initiatives at six-figure annual cloud spend
- Excellent written communication for design documents and platform retrospectives
- Bachelor's or Master's degree in Computer Science, Engineering, or a related field
Preferred Qualifications
- Open-source contributions to ML infrastructure projects
- Conference talks or technical writing on MLOps at scale
- Data engineering at scale (Spark, Dataflow, BigQuery)
- Production GPU operations experience
- Hands-on with FinOps tooling and chargeback models
AI Tools Proficiency
Production engineering at Unico Connect assumes AI tools form part of the daily workflow rather than an experimental augmentation. For this role specifically:
- Claude Code or Cursor for infrastructure and platform code
- LangSmith, Helicone, or Langfuse for production LLM observability
- AI-augmented Grafana, Datadog, or Honeycomb workflows
- Claude or GPT for design document review and post-mortem drafting
What we look for at Unico Connect
Every Unico role expects the same underlying traits — regardless of department or seniority. If these resonate, apply.
Fluent with Claude, ChatGPT, Cursor, Figma AI, or whatever is relevant to your craft. We expect AI tools in the loop, not as a novelty.
Fast cycles, real ownership, low ceremony. You will not be a cog.
Output and outcomes matter more than process. You ship work that moves a metric.
You treat the codebase, the deliverable, and the client relationship as your own.
You joined because you want to ship amazing tech products, not warm a seat.