MLOps Engineer

About the Company

Responsibilities

• ● work will directly determine how reliably tens of thousands of field-service businesses experience intelligent automation.
• ● Run the Linux fleet - provision, harden, patch, configure, back up, and capacity-plan the Linux VM estate across Azure and AWS, and respond to incidents on it.
• ● Own CI/CD - build and maintain the pipelines for a multi-service Python and React estate: build, test, database-migration orchestration, environment promotion, and rollback.
• ● Gate and promote agent versions - run the draft approved deployed pipeline, wire online evaluators in as release gates, and keep one-click rollback through deployment revision tracking.
• ● Operate the multi-provider LLM path - rate limiting, token streaming, per-tenant token budgets, caching, circuit breakers and fallbacks across model providers, with cost and latency metrics flowing into monitoring.
• ● Build the training-to-serving path - stand up GPU training jobs, a model registry, approval-gated deployment jobs, and model serving for the team's machine learning and computer-vision models.
• ● Administer Databricks - workspaces, cluster and usage policies, serverless compute, Unity Catalog, service principals, deployment bundles, and cost attribution.
• ● Build observability - instrument services with OpenTelemetry into Application Insights, run agents or collectors on VMs, and define alerting, SLOs, and error budgets for the agent runtime and voice gateway.
• ● Unify the control plane - bring Azure and AWS hosts under common management for patching, inventory, and compliance baselines enforced through configuration management.
• ● Own identity and secrets - managed identities, workload identity federation for pipelines, secret rotation, least-privilege access, and audit trails.
• ● Own continuity - design backup and restore paths for the data and services the platform depends on, test them on a schedule, and keep documented recovery objectives rather than assumptions.
• ● Own reliability - lead incident response and blameless reviews that end in concrete, tracked follow-ups.

Requirements

• ● + years running production Linux fleets (Ubuntu, RHEL, or equivalent): systemd, storage, networking and firewalls, SSH and key management, TLS and reverse proxies, backups and restores. This is non-negotiable.
• ● Configuration management with Ansible (or equivalent) and patching at fleet scale, plus hardening against recognised benchmarks such as CIS.
• ● Fluent Bash and working Python; comfortable debugging from logs, metrics, and system tools under incident pressure.
• ● Strong Azure experience: container hosting, container registry, Key Vault, service bus, managed PostgreSQL and Redis, virtual networks and private endpoints, VMs and scale sets, managed identities and RBAC.
• ● Infrastructure as code (Bicep, Terraform, or equivalent) and pipeline-as-code with federated, secretless service connections.
• ● Working AWS experience: EC2 fleet management, IAM, VPC, S3, CloudWatch, and managed model access with quotas and guardrails.
• ● Containers and release engineering: multi-service Docker builds, zero-downtime deploys, database-migration orchestration, environment parity, and a tested rollback path. You have built and run a CI/CD pipeline for a team, including the branching model and merge gating.
• ● Observability in practice: OpenTelemetry instrumentation, a metrics and logs backend such as Application Insights or Grafana, and alert design that pages on symptoms rather than causes.
• ● Defining SLIs, SLOs, and error budgets, and running incident response and blameless reviews.
• ● MLOps fundamentals: experiment tracking and a model registry (MLflow or equivalent), orchestrating GPU training, and serving models in production.
• ● LLMOps fundamentals: tracing and online evaluation in production, and LLM serving and routing concerns such as rate limiting, streaming, load balancing, per-tenant budgets, caching, fallbacks, and provider key management.
• ● Security practice: secretless authentication, least privilege, blast-radius thinking, auditability, and dependency and supply-chain hygiene.
• ● Committed to continuous learning, proactive problem-solving, and timely issue identification, with a keen interest in staying current with a fast-moving field.
• ● Strong communicator, experienced in collaborating with cross-functional teams using tools such as Jira and Slack.
• ● Creative and innovative thinker, consistently contributing fresh ideas and solutions in alignment with current technological trends.
• ● Nice to Have
• ● Databricks administration: workspaces, usage policies, Unity Catalog, service principals, deployment bundles, and the Databricks Terraform provider.
• ● Kubernetes and Helm; GitHub Actions.
• ● Self-hosted LLM serving (vLLM, SGLang, TensorRT-LLM) and inference servers such as ONNX Runtime, Triton, KServe, or Ray Serve.
• ● Prometheus and Grafana; load testing of token-streaming endpoints with k6 or Locust.

Benefits

• ● Professional Working environment
• ● Market Competitive Salary
• ● Life Insurance & Medical Insurance (Including Family)
• ● Provident Fund
• ● Gym Facility
• ● Maximum 45 Weekly Hours (Monday-Friday)
• ● Remote Working (During Pandemic Situation)
• ● Company trip
• ● Annual Leaves
• ● Sick & uncapped Compassionate Leaves (As per Company Policy)
• ● Have a chance to work onsite with the UK team