AI Operations Engineer

CatalystOne Solutions AS; Stockholms kommun; Hybrid; SEK 700,000 - 1,100,000; Full time

Benefits offered by this job

Generous time off

Pension and private healthcare

Wellbeing contribution

Laptop and mobile

Home internet allowance

Hybrid work model

Local benefits vary by office

Job summary

CatalystOne Solutions AS is seeking an engineer to ship production AI solutions and build a platform that lets the rest of the company innovate. You’ll design and run agents, automations, and data products used by consultants, support, sales and back-office teams.

You’ll enable others by creating shared components, guardrails, templates and documentation, coaching AI spokes across departments so builds can go production safely and quickly.

Qualifications

• Experience building integrations, backend services, automations or internal platforms.
• Strong foundations in Azure, Microsoft Graph, APIs, serverless, Python or C#.
• Experience delivering AI solutions that real users rely on.
• Good judgement and a habit of choosing the simple solution when it works.
• An understanding of security, reliability and operational ownership.
• Awareness of the trade-offs between capability, speed, complexity and cost.
• The ability to explain technical concepts clearly to non-technical colleagues.
• Experience getting people to change how they work, not just shipping software to them.
• An interest in teaching. You’d rather make ten people capable than be the only one who can do it.

Responsibilities

• Build AI agents, automations and integrations that remove manual work.
• Design the data models, workflows and integrations that make AI solutions reliable.
• Decide where automation fits, where AI adds value, and where human judgement should stay.
• Own what you build: reliability, monitoring, security and cost.
• Work with team leads to change how work is done around what you build, not just hand over a tool.
• Measure whether solutions save time and get used, and act on what you find.
• Turn one-off solutions into reusable patterns, components and templates.
• Set the technical guardrails that let colleagues build with AI safely.
• Coach and review the builders in our AI spoke network, and help their work reach production.
• Run hands-on sessions and contribute to our monthly AI meetups.

Skills

Integrations

Backend services

Automation

Internal platforms

Tools

Azure

Microsoft Graph

APIs

Serverless

Python

C#

Job description

AI is already changing how work gets done at CatalystOne. The next step is for everyone to work that way, not only the people who build it.

We’re looking for an engineer who ships production AI solutions and builds the platform that lets the rest of the company build on it. You’ll write agents, automations and integrations yourself. You’ll also make it easy, safe and normal for colleagues across the business to build their own.

The whole company is your remit. The work is already underway. We’re small enough that something you build can be used by colleagues within weeks.

At CatalystOne, we help organisations create better workplaces through HR technology. Across the Nordics, we combine technology, people expertise and business understanding to help customers improve how work gets done. We now apply the same thinking to our own organisation.

About the AI Hub

The AI Hub is a new function within People & Technology. It exists to identify, prioritise and realise the value of AI across CatalystOne.

We don’t treat AI as only a technology topic. It changes how work is organised, what people need to learn, and where human judgement matters most. The Hub brings process improvement, organisational change and technology together, close to the people doing the work.

The Hub does not do AI for the business. The business owns its processes and results. We provide the technology, the platform and the know-how so teams can move faster on their own.

The role

You have two jobs, and they reinforce each other.

Build. You design, build and run the agents, automations and data products that consultants, support, sales and back-office teams rely on every day. You work directly with the people who have the problem. Some ideas go from a Monday conversation to production the same week. Others need deeper design, data modelling or integration work. You help make those calls, and you own what you build long after launch.

Enable. You turn what you build into something others can reuse. That means shared components, MCP servers, templates, guardrails and clear documentation. You coach the AI spokes we’re growing in each department so they can build safely on their own, and you decide when a colleague’s build is ready for production and when it needs the Hub. Success is measured by what the company ships, not only what you ship.

You’ll work directly with Nikolai Poverud, who leads the AI Hub and partners with the executive team on AI strategy. You lead what happens after a problem is picked: technical design, engineering, operations, the platform, and making sure what we build gets used. You stay close to every solution long after launch.

Already live

This isn’t a greenfield initiative. Today we run:

• An AI support agent that resolves around a fifth of internal IT tickets before they reach the IT team.
• Data products that save consultants time and give us business insight we didn’t have before.
• Automated workflows that cut manual admin and move information between teams.

Your job is to make the next ten of these faster to build, and to make sure more of them are built by the teams themselves.

What you’ll do

• Build AI agents, automations and integrations that remove manual work.
• Design the data models, workflows and integrations that make AI solutions reliable.
• Decide where automation fits, where AI adds value, and where human judgement should stay.
• Own what you build: reliability, monitoring, security and cost.
• Work with team leads to change how work is done around what you build, not just hand over a tool.
• Measure whether solutions save time and get used, and act on what you find.
• Turn one-off solutions into reusable patterns, components and templates.
• Set the technical guardrails that let colleagues build with AI safely.
• Coach and review the builders in our AI spoke network, and help their work reach production.
• Run hands-on sessions and contribute to our monthly AI meetups.

Our stack

Today we use Claude Code and Anthropic models, our own MCP servers on Azure Container Apps, Copilot Studio, Azure AI Foundry, Microsoft Graph, Azure Functions and Logic Apps. We already run agentic workflows in production.

The tools will change. If something better comes along and you can make the case for it, you’ll have room to change the setup. Whatever we run must be secure, understandable, maintainable and cost-effective. It must also be simple enough for others to build on.

What we’re looking for

We care less about your career path than your ability to build things people depend on, and to help others do the same.

You’ll likely bring:

• Experience building integrations, backend services, automations or internal platforms.
• Strong foundations in Azure, Microsoft Graph, APIs, serverless, Python or C#.
• Experience delivering AI solutions that real users rely on.
• Good judgement and a habit of choosing the simple solution when it works.
• An understanding of security, reliability and operational ownership.
• Awareness of the trade-offs between capability, speed, complexity and cost.
• The ability to explain technical concepts clearly to non-technical colleagues.
• Experience getting people to change how they work, not just shipping software to them.
• An interest in teaching. You’d rather make ten people capable than be the only one who can do it.

You don’t need to have run enablement before. You need to want to. You don’t need HR or HCM experience either.

What success looks like

Your first weeks are about learning our infrastructure, the solutions already running, and the colleagues who use them.

You’ll work closely with Nikolai on new opportunities, assessing what is technically possible and where to start.

By day 90, we expect you to:

• Own parts of the existing platform.
• Have contributed to new solutions in production.
• Have shipped at least one reusable pattern or component that someone else has built on.
• Know the spokes and have helped at least one of them get something into production.
• Be able to show usage and time saved for the solutions you own.
• Help build an AI practice from the start.
• See your work used across the company within weeks.
• Shape how a whole company learns to build with AI, not just one team.
• Work directly with the people using what you build.
• Own your work from idea to production and beyond.

As well as your salary, you’ll receive:

• Generous time off: A solid vacation allowance so you can recharge and regroup.
• Financial and health security: A comprehensive pension and insurance package, including private healthcare.
• Your wellbeing, your way: A wellness contribution to help you stay on top of your health.
• Tech tools for efficiency: Laptop, mobile phone with subscription, and a home internet allowance.
• Flexible work setup: A hybrid model that supports your work-life balance, with at least three days a week in the office.
• And more: Local benefits vary slightly by office, and we’re happy to walk you through the details.

CatalystOne does not accept unsolicited resumes from search firms/recruiters. CatalystOne will not pay any fees to search firms/recruiters if a candidate is submitted by a search firm/recruiter unless an agreement has been entered into with respect to the specific open position(s). Search firms/recruiters submitting resumes to CatalystOne on an unsolicited basis shall be deemed to accept this condition, regardless of any other provision to the contrary.

All candidates who receive a job offer will be subject to a background check.