Permanent

Gross annual salary: €37.000 - €48.000

Senior (5 - 8 years)

Requires work permit

Languages: Required: English

WHO WE ARE?
Being the daughter company of Swisscom - leading telecommunications company and one of the most sustainable and innovative companies in Switzerland - we share the same values, vision and goals.As a trustworthy and efficient Swisscom IT service partner, we create valuable customer experiences. We work in strong collaboration with our partners in Switzerland and attract and retain high performers.

We're expanding our footprint in Lisbon and are looking for an Agentic AI Engineer to join our growing AI team.We invite driven professionals to join us, grow with us, and help shape the future.
The team / projectThe Mobile Technology & Voice Value Stream is building an AI tooling suite that automates and supports the operation of Swisscom's mobile and voice network. We are a small, collaborative pod working closely with product owners and business stakeholders in Switzerland, network SMEs and external specialists, and together we're building an internal AI delivery capability for Swisscom Operations.

As Agentic AI Engineer, you are a hands-on individual contributor who designs and builds agentic AI tools that automate operational processes. You help shape architecture decisions, engineering standards, code quality and technical delivery, and help embed AI engineering practices within Swisscom.

What you'll do

• Build agentic workflows: Design and implement agent-based and multi-agent workflows that automate Mobile Technology & Voice operational processes.
• Orchestrate the agents: Develop and evolve orchestrator logic and tool-use patterns for operations automation.
• Ship production-grade LLM apps: Build LLM applications on AWS Bedrock with suitable agent frameworks, and define prompt, context, grounding, citation and response-assembly patterns.
• Make it trustworthy: Implement evaluation for quality, groundedness, latency, reliability and safety, plus guardrails against prompt injection, unsupported claims and unsafe tool usage.
• Keep humans in the loop: Design human-in-the-loop controls and evidence-based decision-support outputs.
• Shape the architecture: Contribute to the end-to-end architecture of the AI tooling suite, take part in design reviews, document decisions in Architecture Decision Records and make pragmatic build-versus-buy recommendations.
• Raise the bar: Uphold the pod's engineering, testing, documentation and code-review standards.
• Deliver as a pod: Plan and coordinate delivery with pod members, remove technical blockers, support onboarding, and pair with external specialists so knowledge stays within Swisscom.
• Turn operations into code: Work with Managers on Duty, network SMEs and operations engineers to convert workflows, SOPs and expert judgement into testable agent behaviour, and demo working software to non-technical stakeholders.

Job requirementsSoftware Engineering

• Strong professional software-engineering background with hands-on delivery responsibility.
• Strong Python skills, with experience designing APIs and distributed services.
• Contribution to technical design, code reviews and engineering quality practices.
  
  

Generative & Agentic AI

• Practical experience building production-grade generative AI or LLM applications.
• Experience with agentic workflows, tool calling, retrieval, grounding and evaluation.
  
  

Cloud

• Practical AWS experience, ideally including Bedrock, IAM, networking, logging and serverless services.
  
  

Soft Skills

• Strong communication skills in English (written and spoken).
• Ability to work autonomously in a distributed, cross-location product team.
• Ability to explain technical trade-offs clearly to non-technical stakeholders.
• Strong ownership mindset combined with a collaborative, pod-first attitude.
• Willingness to share knowledge and support the growth of colleagues.
  
  

Nice to have

• AWS Strands SDK or comparable agent frameworks and runtimes
• Telecommunications, network operations, observability or incident-management knowledge
• Operational AI, explainability, auditability and human-in-the-loop systems
• Collaborating with external delivery partners and supporting knowledge transfer
• Regulated, security-sensitive or enterprise production environments