Agentic AI Engineer
You build AI agents that do real work inside companies: they read information, query systems, make decisions within rules, and hand the decision to a person when needed. You take each agent from prototype to production, with evaluation, observability and cost control.
Apply for this roleAbout us
Vesster implements Agentic AI systems: from the most complex ones, with proven governance models, to early proof-of-concept projects.
Our purpose is to be at the forefront, innovate and lead with AI. We get there through innovation, high standards and resilience.
You will join Vesster's Agentic AI team, which implements AI agents in companies in any sector of activity.
- We build with AI every day, including coding agents in our own development.
- Every project is a chance to apply what AI can already do in real client environments.
- The bar is high, and the commitment is to results in production, not demos.
- We invest in every person who joins. Competence and certification are how we work.
What you will do
- Design and implement agents and multi-agent flows on LLMs, with tools, memory and RAG.
- Integrate agents with client systems through APIs, MCP and A2A.
- Build evaluation sets and measure quality, latency and cost per task before and after every change.
- Put agents into production with logging, traceability, guardrails and human handoff.
- Work on platforms such as Gemini Enterprise Agent Platform, and on open-source stacks when the client requires it.
- Use coding agents in your own work and help the team get the most out of them.
What we are looking for
- Experience building production software in Python and/or C#/.NET.
- Real projects with LLMs — prompting, tool calling, RAG, structured outputs — beyond training exercises.
- A practical grasp of evaluation: you can show whether a change made the agent better or worse.
- Experience with APIs, databases and cloud (GCP, Azure or AWS).
- Fluent English or Portuguese, and the ability to write clearly.
Nice to have
- Experience with agent frameworks (e.g. Google ADK, LangGraph, Semantic Kernel) or with MCP/A2A.
- Experience with agent governance models: observability, continuous evaluation, access control and auditing.
- Knowledge of requirements such as GDPR, the AI Act or sector regulation.
- Cloud or AI certifications (e.g. Google Cloud, Azure, AWS).
Training and certification
Competence and quality are essential requirements. That is why holding certifications is an integral part of the role, not an extra.
- A training plan from day one, tailored to your level and your role.
- Market-recognised certifications (e.g. Google Cloud) are part of the plan — we provide the training and expect you to complete them.
- Time to learn and experiment as part of the job, not only outside it.
What we offer
- Applied AI work in production, for clients in any sector of activity.
- Autonomy and ownership of what you deliver.
- Access to the latest AI tools and models for your work.
Details
- Level
- All levels, from junior to principal. Level is set during the process, not in the ad.
- Location and work model
- All formats — on-site, hybrid or remote — depending on the engagement.
Not an exact match for this profile?
An Agentic AI project needs much more than the agent: API integration, data, data analytics, process analytics, automation, architecture, product management. If you work in one of these areas, or a related one, please apply as well: we want to meet you.
Apply
Apply for this role
A different kind of application: everything you add earns points. Submit whenever you like — a fuller application simply scores higher.
The minimum to submit. Everything after this is you pulling ahead.
Your application score
0 / 419
Tier: Signal · 0 of 25 questions
419 pts still on the table
You can submit at any moment — a more complete application simply scores higher.
