Must-have skills
- 8+ years in software or data engineering, with 3+ years delivering AI/ML workloads on Azure and 2+ years running GenAI solutions in production.
- Strong Python, SQL and PySpark. Writes clean, tested, modular code and knows Git branching and code review well.
- Hands-on with Azure AI Foundry: model deployment, prompt flow, evaluations and managing model catalogues and endpoints.
- Azure Machine Learning: workspaces, compute, pipelines, model registry, managed online and batch endpoints.
- Azure AI Search for vector, hybrid and semantic search, including index design, chunking, embedding strategies and relevance tuning.
- Has designed and built RAG solutions and agentic workflows using orchestration frameworks such as Semantic Kernel, LangChain, LangGraph or AutoGen, with tool and function calling.
- MLOps/LLMOps: automated training and deployment pipelines, prompt and model versioning, offline and online evaluation, and production monitoring for drift, latency, cost, token usage and hallucination rates.
- Containers (Docker) on Azure Kubernetes Service, Azure Container Apps or Azure Functions, with CI/CD through Azure DevOps or GitHub Actions, plus infrastructure as code (Bicep or Terraform).
- Unit, integration and evaluation testing for AI applications, including test harnesses for LLM outputs.
- AI security and responsible AI: Azure AI Content Safety, prompt-injection defences, Managed Identity, Key Vault, private endpoints, RBAC, PII masking and data protection controls.
- Microsoft Fabric (Lakehouse, Data Factory pipelines, OneLake) and Azure integration services (Event Grid, Service Bus, Logic Apps, API Management) for data and AI integration.
- Has led a small engineering team and worked directly with business, data and security stakeholders from discovery through UAT and go-live.
Good to have
Technical
- GitHub Copilot or similar AI-assisted development tools.
- Multi-agent architectures, Model Context Protocol (MCP) and agent observability tools (Azure Monitor, Application Insights, OpenTelemetry, LangSmith or similar).
- Fine-tuning or distilling models (Azure OpenAI fine-tuning, small language models) and managing cost and throughput (PTU vs pay-as-you-go).
- Databricks or Snowflake alongside Azure AI services.
- Knowledge of AI governance frameworks such as the EU AI Act, NIST AI RMF and ISO 42001, or regulated-industry controls (MAS TRM, IM8 for Singapore government).
- Networking for AI workloads: VNet integration, private DNS, hub-and-spoke designs.
Certifications
- Microsoft Certified: Azure AI Engineer Associate (AI-102)
- Microsoft Certified: Azure Data Scientist Associate (DP-100)
- Microsoft Certified: Fabric Data Engineer Associate (DP-700)
- Microsoft Certified: Azure Solutions Architect Expert (AZ-305)
Apply on MyCareersFuture
You apply on MyCareersFuture, where this job was posted. Peter Nowak isn’t the employer and doesn’t check listings.
