Tech Stack
AzureCloudDockerPythonRay
About the role
- Own the end-to-end lifecycle of generative AI features, taking MVPs to reliable, secure, cost-aware production services on Azure.
- Implement features like document summarization and Q&A chat (RAG).
- Build MLOps on Azure, including deployment, monitoring, and observability.
- Establish best practices for evaluation, safeguards, and prompt/response logging with PII-safe redaction.
- Be the hands-on expert and first dedicated full-time AI Engineer, reporting to the IT Lead.
- Influence architecture and future hiring while delivering as a senior individual contributor.
Requirements
- 5+ years in MLOps/Data (SaaS/cloud) with shipped production systems.
- 2+ years of specific experience with Generative AI (LLMs, prompt engineering, retrieval-augmented generation, etc.).
- Proven RAG delivery with production-grade evals (automatic + targeted human review), observability (traces, prompt/response logging with PII-safe redaction), and safeguards (prompt-injection/jailbreak mitigations).
- Python + Docker.
- Experience with MLflow.
- Experience with Azure (Azure OpenAI or compatible) and vector index solutions (Azure AI Search/pgvector/Elastic).
- Experience with async orchestration.
- High ownership; challenges assumptions with data.
- Strong communicator, fluent in English, and thrives in a remote-first, cross-functional environment.
- Experience with Databricks, Ray, or Kedro is a strong plus.
- Must be eligible to work in Portugal (EU citizen or hold a valid Portuguese work permit).