Senior DevSecOps Engineer, AI-First DevOps Change, Developer Platform | 100% Remote

UST; Madrid; A distancia; EUR 90.000 - 120.000; Jornada completa

Descripción de la vacante

UST is seeking a Senior DevSecOps Engineer for a 100% remote role with a UK financial services client. The role focuses on TDD/BDD, DevSecOps, CI/CD, Infrastructure as Code and observability to build secure, scalable pipelines.

You should have experience with C#, Python, TypeScript, Terraform, AWS and Azure, plus GitHub or Azure DevOps. Join a fast-paced AI-first team delivering a developer platform for modern teams.

Formación

• Experience with TDD/BDD, DevSecOps, CI/CD, and infrastructure as code.
• Strong observability and monitoring mindset.
• AI-first mindset with interest in agentic AI tools.

Responsabilidades

• Build rapid, production-ready software in a fast-paced AI-first environment.
• Scale and secure pipelines across large developer toolchains.
• Design and govern autonomous agents and secure token handling.
• Create self-serve dashboards and developer tooling on our platform.
• Ship safe defaults and measure delivery metrics to improve pipelines.

Conocimientos

TDD/BDD

DevSecOps

CI/CD

Infrastructure as Code

Observability

AI-first mindset

GitHub Copilot

Claude

Agentic AI

C#

Python

TypeScript

Terraform

AWS

Azure

GitHub

Azure DevOps

Herramientas

Terraform

AWS

Azure

GitHub

Azure DevOps

GitHub Copilot

Claude

Descripción del empleo

Role Description

We are looking for the very Top Talent…and we would be delighted if you were to join our team!More in details, UST is a multinational company based in North America, certified as a Top Employer company with over 35.000 employees all over the world and presence in more than 30 countries. We are leaders on digital technology services, and we provide large-scale technologic solutions to big companies.

Role Description

We are looking for the very Top Talent…and we would be delighted if you were to join our team!More in details, UST is a multinational company based in North America, certified as a Top Employer company with over 35.000 employees all over the world and presence in more than 30 countries. We are leaders on digital technology services, and we provide large-scale technologic solutions to big companies.

What are we looking for?

We’re looking for a Senior DevSecOps Engineer for a 100% remote position with a leading UK financial services client. Strong experience in TDD/BDD, DevSecOps, CI/CD, Infrastructure as Code and observability is essential. We’re seeking an AI-first mindset, with experience or interest in GitHub Copilot, Claude and agentic AI. Experience with C#, Python, TypeScript, Terraform, AWS, Azure, GitHub or Azure DevOps would be a plus.

The mission

Drop into a fast-paced, AI-first, cross-functional team and contribute by building, ideating and experimenting. We believe first-class management of a developer toolchain will be automated through AI, and we intend to be the team that proves it. From that foundation we build the next generation of products that lift the productivity, security and quality of every team in the organisation. Stated plainly: we are building the best Developer Platform Product team in the industry, and we are hiring the people who will do it.

What You Will Do

Build.

Working software on a real estate, in front of real engineers, fast. Prototypes that earn their place and then get hardened, not documents that describe what we might do.

Put agents to work on the toolchain.

Remediation at fleet scale, pipeline upgrades across thousands of repositories and increasing at new pace, scanner triage, credential rotation, the slow manual work that should not be manual. If an agent can do it, build the agent.

Design How Autonomous Things Are Allowed To Act Here.

Identity, scoping, blast radius, audit trail, stop button, prompt injection. An agent holding a token on a regulated estate is a security design problem, and in this team the person building it is the person who solves it.

Create new products on our own platform.

Self-serve dashboards, MCP servers, skills, developer tooling. Chosen on what customers actually do rather than on what we assume they want.

Ship defaults, not guidance.

We shift left by making the safe, correct path the default in the pipeline people already use. We do not ask developers to do more, earlier.

Experiment with discipline.

Trials are capped and time-boxed, with a decide-or-kill date and a written verdict either way. The discipline is in the killing, not the trialling.

Work test-first, including on the things people usually do not test.

Pipelines, infrastructure, and agents whose output is different every time. Ephemeral environments and contract tests over the golden path, not a plan-diff and a hopeful merge.

Continuous Delivery Practices

Follow continuous delivery practices across all code, platforms and products.

Make the estate observable.

Monitoring, telemetry, logging, tracing and ing across our platform and our pipelines. This is an open area for us today and one of the clearest places to make a mark early, because nothing else here works without it: agents cannot be trusted with work we cannot see, and delivery metrics we cannot measure are opinions.

Talk to customers yourself.

Sit with delivery teams, watch what actually happens, and bring findings back that change the roadmap.

Help Run What We Own.

Everyone here can keep the platform running, and we operate our own platform the way we ask our customers to operate theirs: zero-downtime, defined in code, tested before it ships.

What Good Looks Like

Early on

you are shipping. Something small, real and in production, plus a clear read on where the manual effort in this estate actually goes.

By six months

an agentic flow of your design is doing work that a person used to do, with numbers behind it: what it fixed, how often it was wrong, what it cost. And a second idea in trial.

By twelve months

at least two other engineers can run, debug and extend what you built without you in the path, and you are onto the next thing.

Nothing we operate rests on one person, and that rule applies to your own work from the day you build it.

Essential

What we look for

TDD and BDD, practised to a high level.

Not unit tests written afterwards. Test-first as your default working mode, on production code you were accountable for, sustained under delivery pressure rather than abandoned the first time a date got tight. You believe in these practices, and you can apply them in an agent-first context where the output is different on every run.

A deep understanding of true DevSecOps.

Teams own their pipelines and treat them like their applications and their code. Security is part of building, not a gate at the end of it.

Trunk-based development.

Short-lived branches, continuously integrated, rather than long-lived branches and a painful merge.

Pairing.

It is how we work and how capability spreads, not something we do when there happens to be slack.

Infrastructure as code.

Everything we run is defined in code, and reviewed and tested as code.

Measure, do not assume. Observability as a first-class capability.

Real experience of monitoring, telemetry, logging, tracing and ing, and the judgement to know which signals are worth paying for. You instrument your own work by default rather than waiting for an incident to tell you what you should have been watching.

Delivery metrics driven to the elite DORA band.

Release on demand, lead time and time to restore service inside the hour, change failure rate in the low single digits. You have carried a team or a platform toward numbers like those, and you know the difference between improving the metric and improving the system that produces it.

AI-first, and already on the journey.

We use Claude and GitHub Copilot alongside a growing fleet of cloud agentic tools to do our work, and we are deep in the practice of working out what engineering looks like in an AI-first world. You are on that journey now and keen to keep exploring it at pace.

Experience And Interest, Rather Than Mastery

CI/CD platforms, patterns and practices. Some real experience, and genuine interest in where they are going next.

Nice to have

Languages:

C#, Python, TypeScript

Infrastructure:

Terraform, AWS, Azure

Toolchain:

GitHub, Azure DevOps, Artifactory

What We Are Actually Reading For

You build things.

Your instinct under uncertainty is a working prototype, not a proposal.

Product instinct.

You ask who the user is and what they will actually adopt before you ask what is technically interesting.

You change your mind on evidence

and you can kill your own work when the measurement says so, without needing to be talked into it.

You teach it out.

Pairing, writing, speaking internally. Capability that stays in your head is not capability we have.

Strong engineering fundamentals in whatever you work in, because all of this is still software.

What We Are Not Asking For

• A set number of years in anything. The essentials above are about practice you can demonstrate, not about how long you have been doing it.
• A machine learning research background. There is no model training in this role.
• Prior financial services experience.
• Depth in every tool we run. We run a lot of it, and nobody arrives knowing all of it.

How We Assess

• Hands-on pairing, using AI tools to solve a real problem

Not a whiteboard algorithm. You pair with one of our engineers on something