AI Platform Engineers (AI/GenAI | Cloud | Kubernetes) - Contract - Onsite - Sandton

ENVIRONMENT:

Our client is seeking highly specialised AI Platform Engineers to design, build, operate and optimise enterprise-grade AI infrastructure within a complex, regulated environment. This is not a general cloud engineering, IT infrastructure or data science role. The successful candidates must have hands-on experience supporting production AI workloads across multi-cloud environments and must be comfortable working across AI infrastructure, model serving, agentic AI, security, observability, infrastructure-as-code and AI cost governance.

DUTIES:

· Design, deploy and optimise scalable multi-cloud AI platform infrastructure.

· Build reusable platform components for AI gateways, model serving, vector databases, data pipelines and GPU workloads.

· Develop and maintain infrastructure-as-code using Terraform, Pulumi, CloudFormation or equivalent technologies.

· Design infrastructure supporting agentic AI, including orchestration environments, tool-calling, agent memory, state management and multi-agent communication.

· Implement cloud-agnostic model-serving patterns that support workload portability.

· Define and manage AI platform SLAs covering availability, inference latency, throughput and reliability.

· Implement platform observability, monitoring, incident management, release management and operational runbooks.

· Design and implement zero-trust security controls for AI platforms.

· Manage AI compute expenditure through cost attribution, chargeback/showback, workload optimisation and usage reporting.

· Maintain technical documentation, architectural decision records and governance evidence.

· Mentor engineers and contribute to platform engineering standards and delivery practices.

REQUIREMENTS:

• Senior level: approximately 5–8 years of relevant cloud and AI platform engineering experience.

• Lead/Principal level: approximately 8–12 years of relevant experience, including technical leadership and responsibility for engineering teams or platform squads.

Mandatory Technical Experience:

• Candidates must demonstrate meaningful production experience in most of the following:

• At least two of the following AI ecosystems:

o AWS Bedrock or SageMaker

o Microsoft Azure AI Foundry or Azure OpenAI

o Databricks AI

o Enterprise Hugging Face deployments

• Kubernetes, Docker, Helm and containerised platform services.

• Terraform, Pulumi, AWS CDK, CloudFormation or equivalent infrastructure-as-code.

• CI/CD and automated deployment of cloud or AI platform components.

• Production model-serving infrastructure, AI gateways or inference endpoints.

• Platform observability using tools such as Prometheus, Grafana, Datadog, OpenTelemetry or Databricks Lakehouse Monitoring.

• Cloud security, identity and access management, including OAuth/OIDC, JWT, RBAC or ABAC.

• Production incident management, SLAs, release management and operational readiness.

• Experience within banking, financial services or another highly regulated enterprise environment.

Specialist AI Experience:

• Candidates should demonstrate practical experience in one or more of the following:

• Agent orchestration frameworks such as LangGraph, AutoGen, AWS Bedrock Agents or Microsoft Foundry Agent Service.

• Model Context Protocol, tool-calling APIs and agent state or memory management.

• Retrieval-augmented generation and vector database infrastructure.

• Cloud-agnostic model serving using tools such as ONNX, BentoML, Triton Inference Server or vLLM.

• MLOps platforms such as MLflow, Kubeflow or Airflow.

• GPU cluster management and inference or training workload optimisation.

• Prompt-injection prevention, output filtering, data-exfiltration controls and AI threat modelling.

•

AI Finops Experience:

• Candidates should have experience with some combination of:

• AI or cloud cost attribution and tagging.

• Chargeback and showback models.

• Token, GPU, DBU or provisioned-throughput cost management.

• Rightsizing, workload scheduling and reserved or spot-instance optimisation.

• Cost dashboards, anomaly detection and cost-per-use-case reporting.

• Communicating technical cost trade-offs to senior technology, business or finance stakeholders.

Qualifications:

• Postgraduate qualification in Computer Science, Information Technology, Data Science, Mathematics, Statistics, Engineering or a related quantitative field.

• A Master’s degree is preferred and may be required for certain senior appointments.

• Relevant certifications are strongly preferred, including:

o AWS Solutions Architect Professional or AWS Machine Learning

o Microsoft Azure AI Engineer

o FinOps Certified Practitioner

o Certified Cloud Security Professional or equivalent

o HashiCorp Terraform Associate

Kubernetes certification