Senior AI Platform Engineer

Aquablu B.V

Amsterdam

On-site

EUR 76,000 - 100,000

Full time

14 days+

Get more replies from employers

Send a job-specific resume in minutes.

Benefits offered by this job

Office in Amsterdam (Houthavens)

Equity via Aquablu for All

Healthy lunches

Urban Sports Club membership

Travel compensation

Pension plan

Quarterly team events

Job summary

Aquablu B.V. is seeking a deeply hands-on, senior platform engineer to own the core infrastructure that powers rapid, safe product delivery. You will shape the prototyping environment, CI/CD pipelines, test harnesses, AI tooling, and the runtime for AI agents and automated workflows.

This is a high-ownership role with broad autonomy and responsibility for quality across the digital org. Based in Amsterdam, you’ll lead the platform strategy, set testing and evaluation standards, and ensure fast,

Qualifications

Hands-on production systems experience.
Senior-level ownership of a system or platform end to end.
Experience with agentic systems in production: LLM workflows, agent orchestration, or automation.
AI-native coding mindset and daily use of AI agents.
Platform/infrastructure background and developer experience.
Ability to unblock the team and ship with speed while maintaining quality.
Test automation that runs without constant supervision.
Eval/reliability tooling design for AI/LLM systems.
Ability to think in eval suites, guardrails, and pipelines.
Adaptability to new tools and rapid shipping under uncertainty.
No title limiting; focus on outcomes and impact.

Responsibilities

Build the sandbox and scaffolding for rapid idea testing by builders.
Own CI/CD and deploy infrastructure for fast, repeatable, safe shipping.
Create automated test & quality gates integrated into CI pipelines.
Own QA, testing and evaluation strategies for the digital org.
Develop evaluation infrastructure with automated eval suites and guardrails.
Maintain runtime for AI agents and automated workflows with proper versioning and cost control.
Build AI tooling rails to empower builders to use AI effectively.
Provide observability signals to monitor production health across products and agents.
Shape tooling strategy across the team to minimize chaos and bottlenecks.

Skills

Hands-on engineering

Senior ownership

Agentic systems

AI-native coding

Platform instinct

Fast learner

Test automation

Eval & reliability tooling

Guardrails & pipelines

Decision making under uncertainty

Shipping speed with quality

Job description

Who are we?

At Aquablu we’re disrupting the marketplace with a revolutionary (flavored) water system. Brought to life in 2018, our founders Marnix and Marc promised a healthy but tasty alternative to soft drinks. Today, our water is tapped in almost 16 countries, and we continue to make waves by teaming up with industry leaders like Heineken, Adyen and Marriott Hotels. Join us in our mission to save one billion plastic bottles by 2030!

“That might sound really ambitious, but our team is part of an inspiring movement in an industry that’s thirsty for change.”

- Marc, Co-founder of Aquablu

About the Job

How fast our product builders go from idea to something real in users' hands is how our digital products win. You build the engine that makes that speed possible (and safe) by default.

You build the platform underneath everyone who ships: the prototyping environment, the deploy pipelines, the test and eval harnesses, the AI tooling, and the runtime our AI agents and automated workflows run on. When a builder has an idea on Monday and it's in front of users on Friday, your platform is why — and why everyone can trust it works. Every hour you save the team compounds across every product we build.

This is a deeply hands‑on senior engineering role with wide autonomy. You scope, decide, and deliver the platform yourself — you're the domain expert everyone who ships comes to, and the standards you set become how the whole team works. You're also the quality owner for the digital org: the testing strategy, the eval strategy, and the bar itself sit with you, encoded in infrastructure rather than enforced by meetings. You're in the code every day, building real infrastructure with AI as your primary collaborator. One day you're cutting a deploy pipeline from hours to minutes; the next you're building the eval harness so a builder can trust a feature behaves before it ever reaches a user; the day after you're putting guardrails and monitoring around an agent that runs company workflows unsupervised.

You own the rails; the builders and engineers own what runs on them. Your job is that everything on those rails ships fast, behaves predictably, and stays affordable.

If building the thing that makes a whole team faster excites you more than shipping any single feature, this is the role for you.

This role reports into our Head of Digital Product and is part of Aquablu's Software & Intelligence Engineering team.

It’s a full‑time position in our office in Amsterdam, Houthavens and office presence is 4 days minimum per week.

What You'll Build & Own

Prototyping & discovery environment: Build the sandbox and scaffolding that lets builders test an idea in hours, not weeks.
CI/CD & deploy infrastructure: Own the path to production — fast, repeatable, and safe enough that shipping doesn't need permission.
Automated testing & quality gates: Build the test infrastructure that proves code works before it ships — AI‑generated and AI‑maintained test suites, regression coverage, and quality gates wired into CI so every change gets checked by the pipeline. You define what "good enough to ship" means and encode it, so quality holds at speed without a human standing in the way.
QA ownership: You're the quality owner for the digital org. The testing strategy, the eval strategy, and the bar itself sit with you — encoded in infrastructure, not enforced by meetings.
Evaluation infrastructure: Make reliability measurable on systems where the same input can behave differently each run. Ship this as automated eval suites and guardrails builders can trust before a feature reaches users.
Agent & workflow infrastructure: Own the runtime our AI agents and automated workflows run on — orchestration, permissions, versioning, cost, and monitoring. Agents that do real work are production software; you make them behave like it.
AI tooling & enablement: Build the rails that let every builder use AI well — and keep raising the bar as the tools change.
Observability & reliability: Give builders the signal to know what's working in production without going looking for it — for products and for agents alike.
Tooling strategy: Evaluate, adopt, and roll out new tools across the team without chaos. You anticipate what the team needs before it becomes a bottleneck.

What You Bring

Real, hands‑on engineering experience: You've built and shipped production systems yourself. This is a builder's role, not an architect‑on‑paper or manager role — you live in the code.
Senior‑level ownership: You've operated at senior level before: owning a system or platform end to end, making the calls in ambiguous or high‑stakes situations, and being the person others came to for judgment — not just execution.
Agentic systems expertise: You've built or operated agentic systems in production: LLM workflows, agent orchestration, or automation that runs unsupervised — and you know what it takes to keep them reliable, safe, and affordable.
AI‑native coding mindset: AI‑native coding is how you work, not a skill you're adding. You build with AI agents every day, you have a sharp setup, and you push the boundary of what that lets you ship. If your last six months of work doesn't look meaningfully different because of AI, this isn't your role.
Platform instinct: Background in platform, infrastructure, developer experience, or ML/AI tooling — and a track record of building things other engineers actually adopted.
Multiplier mindset: You’d rather unblock the team than be the bottleneck the team waits on.
Test automation proficiency: You've built test and quality automation that runs without you in the loop. You get the pipeline and AI agents to generate, run, and maintain the checks, and you have the judgment to set the coverage and quality bar that keeps speed safe.
Eval and reliability tooling: Experience building eval or reliability tooling for AI/LLM systems.
Thought process: You think in eval suites, guardrails, and pipelines, not gatekeeping and review queues.
Fast learning and shipping under uncertainty: You've adopted tools in the last six months that didn't exist a year ago.
No title limiting: We don't care what your last title was. We care what you've made — and what you've made other people able to do.