AI Engineer, Internal AI - Typescript

Start: ASAP or in 1 months time Duration: 6 months Stockholm, on-site The team The Internal AI team is responsible for enabling AI-native ways of working across the organisation. The team works across several areas in parallel: building internal AI agents and tools, enabling teams to develop their own solutions, and creating the standards, access models and guardrails that allow AI to be used effectively and securely at scale. The team owns both the underlying AI platforms and the internal products built on top of them. This includes internal knowledge agents, AI-powered workflow automation, structured and agent-accessible knowledge bases, shared infrastructure for internally developed AI applications, and reusable AI capabilities that can be adopted across the organisation. The team is small, but its mandate spans the entire organisation. This means combining hands-on engineering with platform thinking, product ownership and close collaboration with different parts of the business. The challenge This is an organisation where AI adoption is already high and users are comfortable experimenting with the latest tools. The challenge is therefore not convincing people to use AI, but building solutions that are genuinely valuable, reliable and better than what individual teams could easily create themselves. At the same time, the organisation needs common foundations that individual teams cannot efficiently build on their own. This includes infrastructure, access and permission models, knowledge architecture, monitoring, evaluation and appropriate guardrails. There will not always be a detailed specification. Part of the role is understanding the underlying problem, determining what should be built and creating scalable solutions that work across the organisation. What you will do

• Build internal AI agents, tools and platforms and treat them as long-term products rather than one-off projects.
• Take broad business goals, determine what is actually required and translate them into concrete technical solutions.
• Own solutions end to end, from architecture and implementation through production, adoption and continuous improvement.
• Help shape what should be built, how it should be prioritised and how the overall internal AI ecosystem should evolve.
• Design foundations including knowledge architecture, access and permissions, evaluation, monitoring and risk-based guardrails.
• Work directly with users and stakeholders across the organisation, including teams outside engineering.
• Ensure internal AI solutions are scalable, secure, reliable and genuinely useful to the people using them.

What we are looking for

• Strong engineering fundamentals. You build robust, production-ready systems and take responsibility for delivering solutions that work in real-world environments.
• Deep experience with LLM-based systems. You have hands-on experience with agents, retrieval, evaluation and tool use. You understand where these systems can fail and know how to design for reliability and performance.
• A strong sense of ownership. You don't wait for perfect specifications. When requirements are unclear or could be improved, you challenge assumptions and propose better solutions.
• Technical, domain and business understanding. You can discuss architecture in depth while also understanding the underlying business needs and connecting the two.
• Openness and collaboration. You welcome feedback, build on other people's ideas and are comfortable changing direction when new evidence emerges.
• Comfort with ambiguity. You enjoy working in fast-moving environments where teams evolve, priorities change and decisions are revisited as new information becomes available.
• Strong TypeScript experience. You are comfortable building and maintaining modern production systems using TypeScript.
• Comfort with infrastructure. You are willing to work across the stack, including infrastructure, Terraform and cloud environments, rather than treating infrastructure as someone else's responsibility.
• AI-native development. You use AI coding agents as a natural part of your daily development workflow, not simply as an occasional experiment.