Hybrid (Lisbon, Braga or Coimbra)

What we're looking for

A DevOps Engineer to join a international project.

Working model: Hybrid

English level: B2 (mandatory)

Candidates must be living in Portugal

Main Responsibilities

Solution Takeover & Knowledge Transfer

• Lead structured handover of AI solutions delivered by external vendors, across any in-scope solution type (agents, copilots, RAG apps, document intelligence services, integrations, fine-tuned LLMs, etc.).

• Review code, prompts, configurations, architecture, and documentation; identify technical debt and define a remediation backlog.

• Produce or refresh runbooks, SOPs, prompt and model documentation, and operational documentation for solutions taken into run mode.

• In-House AI Solution Development

• Design, develop, and ship AI features and full solutions using authorized AI stack, selecting the most appropriate pattern for each business problem (rule-based automation,RAG, agentic workflows, fine-tuning, multimodal AI services, etc.).

• Build solutions with Azure OpenAI, Microsoft Copilot Studio, M365 Copilot extensibility, Azure AI Foundry, Azure AI Services (vision, speech, translation, document intelligence), and custom Python/TypeScript services as appropriate.

• Apply enterprise patterns for grounding, evaluation, guardrails, and responsible AI consistent with AI governance and Microsoft Copilot Studio governance frameworks.

• Contribute reusable components (connectors, prompt templates, agent skills, evaluation harnesses) to the AI platform library.

• Collaborate with the AI Solution Lead on architecture reviews, pattern selection, and design validation.

DevOps & Platform Operations (incl. LLMOps)

• Build and maintain CI/CD pipelines (GitHub Actions / Azure DevOps) for AI solutions, following GitOps and IaC standards.

• Apply LLMOps practices: prompt versioning, evaluation datasets, A/B testing of prompts and configurations, automated quality evaluation, and regression testing.

• Implement observability (logs, traces, cost, usage, response quality, grounding quality) and act on alerts.

• Manage application lifecycle, secrets, environment promotion, and release management.

• Optimize consumption costs (tokens, compute, storage, inference calls) and report on unit economics per solution.

User Support (L3)

• Act as the L3 escalation point for in-scope AI solutions; partner with L1/L2 team per the existing service model.

• Investigate and resolve incidents, problems, and quality issues (response quality, hallucinations, grounding failures, integration errors); drive permanent fixes back into the codebase, prompts, or configurations.

• Maintain a knowledge base of known issues, workarounds, and FAQs for end users and support partners.

Backlog Management & Business Engagement

• Run discovery and refinement sessions with business product owners; translate user needs into well-formed user stories with acceptance criteria.

• Triage, estimate, and prioritize the backlog together with the business; plan iterations and communicate releases.

• Demo increments, collect feedback, and steward adoption together with business champions.

GitHub Copilot Enablement (Secondary)

• Act as a subject-matter expert for GitHub Copilot, authorized AI coding platform.

• Provide best-practice guidance, internal enablement sessions, and prompt/usage patterns for other ADS developers.

• Monitor adoption signals and feed insights back to the AI governance team.

Experience

• 5–8 years of professional software development experience.

• At least 2 years building or operating generative AI / LLM-based solutions in production.

• Demonstrated breadth across multiple GenAI solution types (e.g., conversational, agentic, RAG, document intelligence, integrations).

• Proven track record of taking over and stabilizing vendor-built code bases.

• Hands-on experience with backlog management and direct business stakeholder engagement.

Technical Skills (Must-have)

• Strong Python and/or TypeScript; solid software engineering fundamentals (testing, version control, code review).

• Azure cloud (App Service / Container Apps / Functions / AKS), Azure OpenAI, Azure AI Foundry, Microsoft Graph, REST API integration.

• Practical experience with at least three GenAI solution patterns from the following: conversational agents, agentic/multi-agent workflows, RAG, document intelligence, AI-driven integrations.

• Prompt engineering, evaluation, and guardrail design for LLM-based applications.

• CI/CD pipelines, IaC (Bicep / Terraform), containerization (Docker), observability tooling.

• Microsoft 365 ecosystem awareness (Copilot Studio, SharePoint, Teams as data sources and surfaces).

Technical Skills (Nice-to-have)

• Agentic frameworks (Microsoft Agent Framework, LangGraph, Pydantic AI etc.).

• Vector stores and knowledge-graph tooling (Azure AI Search, Qdrant, Neo4j, LlamaIndex etc.).

• Azure AI Services (Document Intelligence, Vision, Speech, Translator) for multimodal application development.

• Fine-tuning and model customization on Azure OpenAI / Azure AI Foundry.

• GitHub Copilot enablement experience.

• Azure AI Engineer (AI-102) certification.

Technical Skills (Must-have)

• Strong Python and/or TypeScript; solid software engineering fundamentals (testing, version control, code review).

• Azure cloud (App Service / Container Apps / Functions / AKS), Azure OpenAI, Azure AI Foundry, Microsoft Graph, REST API integration.

• Practical experience with at least three GenAI solution patterns from the following: conversational agents, agentic/multi-agent workflows, RAG, document intelligence, AI-driven integrations.

• Prompt engineering, evaluation, and guardrail design for LLM-based applications.

• CI/CD pipelines, IaC (Bicep / Terraform), containerization (Docker), observability tooling.

• Microsoft 365 ecosystem awareness (Copilot Studio, SharePoint, Teams as data sources and surfaces).

Technical Skills (Nice-to-have)

• Agentic frameworks (Microsoft Agent Framework, LangGraph, Pydantic AI etc.).

• Vector stores and knowledge-graph tooling (Azure AI Search, Qdrant, Neo4j, LlamaIndex etc.).

• Azure AI Services (Document Intelligence, Vision, Speech, Translator) for multimodal application development.

• Fine-tuning and model customization on Azure OpenAI / Azure AI Foundry.

• GitHub Copilot enablement experience.

• Azure AI Engineer (AI-102) certification.

Education

• Bachelor's or Master's degree in Computer Science, Software Engineering, or equivalent.

Behavioral Competencies

• Outcome-oriented, pragmatic, comfortable balancing build versus run.

• Strong written and verbal English communication; able to translate between business and technical audiences.

• Continuous improvement mindset; operates well in a hybrid global team with vendors and internal stakeholders.

• Curiosity and self-learning attitude in a rapidly evolving AI landscape; able to pick the right tool/pattern for each problem rather than defaulting to one approach