Role Overview
Wert is crypto-to-fiat payment infrastructure: our widget lets people buy crypto with a card inside someone else's product. Real money, real regulators, production volume.
We are hiring our first in-house DevOps engineer. Our infrastructure has been run by an external provider for three years, and we are bringing that function inside.
Be clear about the shape of this job. You will be the only DevOps engineer at Wert, and the work is not to write a roadmap and hand it to a team — there is no In House DevOps team (yet). It is to take a live payment platform into our own hands, with yours, while it keeps serving traffic. You get a part-time internal partner who knows the current setup, and support for the decisions you make.
If you are looking for a lead or architect seat where you set direction and others execute, this is not that role. If you are at your best inside a system rather than above it — and you like being the person who actually owns it — read on.
What you will own
Bringing infrastructure in-house
- Take over a platform you did not build: map it, absorb what is undocumented, take ownership of access, and move active infrastructure work inside the company.
- Decide what we keep from the current stack and what we replace — then execute that decision yourself rather than delegating it. We would rather you learn and optimise what exists before proposing an expensive migration away from it.
- Put infrastructure knowledge somewhere Wert can read it. Today too much of it lives in heads.
CI/CD and developer experience
- Own the delivery pipeline end to end and make it something engineers trust: fast, predictable, and strict where strictness matters.
- Build per-PR review environments that developers and QA drive themselves, wired for automated test gating — including routing external provider callbacks into ephemeral environments.
- Root-cause the failures instead of restarting the job. This is the difference we hire for.
Platform and infrastructure as code
- Architect and maintain the environment in Terraform (or Pulumi/CDK), with an internal library that lets product engineers self-serve the simple things without waiting on you.
- Own Kubernetes: workload configuration, resource behaviour, upgrades, and the boring reliability work that keeps a payment platform boring.
Running a payment platform
- Observability with thresholds you can defend. Metrics, logs, and alerts that fire on what matters and stay quiet otherwise.
- Third-party failure is our normal state. Blockchain nodes, KYC providers, and payment providers degrade without telling us. Detecting that, telling it apart from our own failures, and making it visible to whoever is on duty is core to this role, not an edge case.
- Regulated environment. We operate under European payment and crypto-asset regulation. Practically: your processes — access control, change management, backup and restore — need to leave a documented, evidenced trail, not just work.
- Cost ownership. You own the cloud bill: visibility first, then efficiency, then keeping new features from being built expensive. We would rather hear that a saving isn't there than get a number invented to please us.
Working with engineering
- You sit in the development process from design onward, not behind a ticket queue.
- You teach developers and QA how the pipelines work and how to use the tools you build. Not a management track - the influence here comes from being the person who knows this best.
Requirements
- Migration experience - mandatory. You have personally moved infrastructure off a managed, legacy, or externally-run platform. Not "was on a team where it happened": your hands, your changes, your rollbacks.
- Real operational experience. You have carried on-call, chosen alert thresholds yourself, and been woken up by them.
- Cloud: expert-level AWS — EC2, S3, Aurora, IAM, CloudWatch.
- Orchestration: deep Kubernetes, vanilla and/or distributions such as Deckhouse.
- IaC: Terraform (Pulumi/CDK welcome); comfortable reading and writing Go.
- CI/CD: Jenkins and GitHub Actions — pipeline design, not just usage.
- Observability: Prometheus, Grafana, OpenSearch, Sentry, and log analytics at volume.
- Security: secret management with HashiCorp Vault.
- Databases: working knowledge of PostgreSQL/Aurora and ClickHouse — enough to support and optimise, not to own the data model.
What we look for in a person
- Operator, not narrator. We probe for what you did with your own hands. "I don't know" is a free and fully acceptable answer here; the expensive answer is a confident one that collapses on the second question.
- Pragmatic before ambitious. The best first move in an inherited system is usually to understand it, not to replace it.
- Straight back with vendors. When someone says a thing is impossible, you separate the technical claim from the commercial one and ask for the reasoning.
- Root cause over band-aid, even when the band-aid ships tonight.
Working at Wert
- Remote, on a GMT+2 working day. The team runs on GMT+2 (Central/Eastern European time) and we ask that your day lines up with it — this role is on the hook when things break, so overlap is the point, not a formality. Where you live is otherwise up to you. ~40 people.
- Compensation: a defined band for the role, discussed openly in the first call. We would much rather talk about the number early than late.
- Autonomy with backing. You own your area and make the calls in it. Being the only DevOps engineer here means ownership, not isolation.
- Work that is visibly load-bearing. When your infrastructure is good, people can buy crypto in products all over the internet. When it isn't, everyone finds out. Few places make the impact of this role that legible.
