Agentic Engineer (Medior)

Build, ship and run software with AI, as part of a team

Department Product Engineering · Role Agentic Engineer (Medior) · Location Amsterdam · Work arrangement Hybrid (50%) · Hours 0.8 to 1.0 FTE

Software that works in production, built with AI and owned by you

Do you see AI as a way to get more done, but know that the output still has to be correct? Therapieland is looking for an Agentic Engineer who builds, ships and runs software for our platform as part of a FLOW team, with AI doing a large share of the execution. You work across the stack and have depth in at least one area. You understand the problem, give AI the right context, and verify the result before it reaches a client or professional.

About Therapieland

Mental healthcare is under pressure. Demand keeps growing, and waiting times are the result. At Therapieland we are working on a different outcome: our digital solutions help organisations organise value-driven mental care around the person. Every day we support professionals and their clients in mental health care (GGZ) and the social domain. We develop inspiring, intelligent and connecting psychology on demand. Samen mentaal veerkrachtig (together, mentally resilient).

As part of Visma, we embrace new technology, AI first, to make mental healthcare smarter, more personal and more scalable.

The role of the Agentic Engineer

Therapieland builds in FLOW: small teams that take a problem from discovery to a measured outcome in production, with as few handovers as possible. A FLOW team brings together a product manager, a product designer and engineers, and works closely with customer success, marketing, the helpdesk, and the clients and professionals who use our platform.

AI now produces much of the code, tests and infrastructure that engineers used to write by hand. That changes what one engineer can deliver. It does not change the need to understand the problem, choose the right approach and know whether the result is correct. That is where you spend most of your time. Good input produces good output.

As a medior you work independently. You own standard solutions end to end, help own and improve the team's agentic tooling, and stay accountable for what you build. Our platform is regulated e-health, so speed never overrides our security, privacy and clinical safety standards, including ISO 27001 and NEN 7510. AI executes; people remain accountable.

Your responsibilities

Deliver outcomes with your team: you help shape solutions from discovery onward, build them, ship them, and stay involved until their effect is measured. You own regular changes from acceptance criteria to monitoring in production.
Build with AI, and improve how the team does it: you use our internal agentic tooling daily and, together with other engineers, maintain and improve its workflows, skills, context and MCP connections. You create new workflows where the team needs them.
Own quality from the start: you define acceptance criteria before building, challenge the context you give AI, and recognise output that looks correct but is subtly wrong. You build verification, harnesses and evals into team workflows.
Run what you build: you make sure your systems are monitored on the shared tooling and stay healthy. When something breaks in a system you own, you fix it and use what you learned to improve the design.
Work across the stack, with depth: you work in at least three of front-end, mobile, backend, data and APIs, and infrastructure, and you build depth in one area: quality engineering, cloud and infrastructure (Google Cloud), front-end and mobile, or backend and data.
Make good decisions, and help others make them: you propose designs and record decisions. When someone else has to decide, you bring two or three options with their pros and cons and your preference. Architecture questions go to the weekly architecture task force, with a proposal.

Who are you?

Engineering experience across the stack: at least three years as a software engineer, with production work in at least three layers of the stack. You have a higher professional education, preferably in a software-related field.
An AI-native way of working: you use agentic tooling every day, know where it fails, and can give examples of errors you caught in AI output. You can explain every change you commit.
Analytical and critical thinking: you investigate problems independently and evaluate code, tooling and test results objectively.
Collaboration and feedback: you pair and spar with colleagues, give direct and constructive feedback, and ask for it in return. You align early on dependencies.
Clear communication: you explain to product managers, designers and colleagues what is and is not possible, and come with alternatives rather than only a no. You write clearly, because context and decisions are written down. Excellent command of English; Dutch is a plus.
Security and privacy by design: you apply security and privacy in your own solutions and recognise when a change needs extra review.
A flow mindset: you measure your contribution by outcomes in production, not by tickets closed, and you ask why a dependency exists before managing it.

Preferred: experience building agentic workflows, skills, harnesses or evals; Google Cloud (Cloud Run) and Terraform; native mobile or the Vuetify design system; test automation; experience in healthcare or another regulated domain.

What makes this role distinctive

You build in a team, not alone with AI. You pair with other engineers and work with a product manager, a product designer and stakeholders inside and outside Therapieland. You share responsibility for the outcome, and feedback is a normal part of how we work.

A clear growth path. The role has three levels (junior, medior, senior) with named depth areas and explicit criteria for moving up. Depth and specialisms count alongside breadth.

The Human Turn in engineering. Therapieland continuously asks where human presence changes the outcome in ways that matter. As an Agentic Engineer you work with that question every day: what needs human judgement, what can be delegated to AI, and how you remain answerable for the result, whoever or whatever executed it.

What we offer

A hybrid work environment (50% office)
A competitive salary of €3,263 to €5,098 per month (Baarda Expert, medior), based on 40 hours per week
8% statutory holiday allowance and 30 holidays based on full-time employment
Always off on 5 May, Liberation Day
Travel reimbursement (NS Business Card or mileage)
Lease-a-bike programme
Access to the best agentic engineering tools and platforms. We do not cap tooling investment for engineers working in FLOW
Room for personal development and for trying new tools and methods
A team of professionals who work from shared values

We are committed to building a diverse and inclusive team. Good engineers come from many different backgrounds and career paths. If you are motivated by making AI-driven mental healthcare safe and effective, we want to hear from you.

Agentic Engineer

Begin met je e-mail