AI Integration Engineer

About the role We are setting up a new team at Frends that works a bit differently from the rest of the organization. It is small and independent, part of our product organization, with room to set its own pace and ways of working. Its job is to make migrating to Frends from other integration platforms fast, predictable and repeatable, using AI tooling we build ourselves. We work close to customer projects. When a new challenge comes along, such as a source platform we haven't migrated from before, the team takes it on first: building the tooling, solving the problems as they come up, and doing the Frends development. Once the approach is proven, we hand it over to our Delivery teams and partners and move on to the next one. This means you will often start without a ready playbook. You will be trusted to figure things out, try different approaches, and learn fast, with the support of a team. We are looking for two AI Integration Engineers (mid or senior level) to join us. What you'll do: Build the AI tooling (about 40%) Build and maintain the AI tooling (agent skills, scripts, prompts, automated tests) that consultants, partners and customers use to analyze, estimate, design and build Frends integrations. Add support for new source platforms. Own the quality of the tooling: regression tests, versioning and release notes. Deliver migration projects (about 40%) Do hands-on Frends development on migration projects, especially the first projects on a new source platform. Analyze customers' existing integrations, then estimate the work and design the Frends solution. Join customer technical sessions, mostly remotely. Close the loop (about 10%) Turn what goes wrong on projects into backlog items and fixes in the tooling. Track how accurate the estimates were against actual effort. Take part in developing new tools and services. Get others to use it (about 10%) Coach consultants and partners on the tooling and pair with them on their first migrations. Own the playbooks and documentation that let a platform be handed over to Delivery. Contribute to turning the tooling into a service for partners and customers. What we're looking for: You have around 5+ years in software or integration development, with strong experience in at least two of the following: Enterprise integration: hands-on development on an enterprise integration platform (iPaaS or ESB), such as Frends, Azure Integration Services, MuleSoft, BizTalk, Boomi, webMethods, TIBCO or SAP PI/PO. AI-assisted development: building with AI coding agents in your daily work (for example agent workflows, skills or MCP), together with a backend or integration development background. Hands-on development: you write code and build solutions yourself. Customer work: experience working directly with customers on technical topics, for example workshops, requirements gathering or technical solution discussions. In addition, you are a fast learner who thrives in a fast-paced environment, is curious about solving problems, and is willing to jump into unknowns. You also bring: Hands-on use of AI coding agents (for example Claude Code, Cursor or Codex) in real development work, beyond using them as chat assistants. A solid understanding of APIs, data formats and messaging (for example REST, SOAP, JSON, XML, queues). Programming skills in at least one general-purpose language (for example C#, Python, TypeScript or Java). The ability to read and understand unfamiliar code and integration logic quickly. Version control and modern development practices (Git, code review, CI/CD). Fluent English, spoken and written, with developers and with customers. Nice to have: Experience in migration projects from one integration platform to another, or with several integration platforms and how the same concepts map between them. Building tools, automations, or internal platforms that other developers use. Validating and testing AI-generated code or output, for example evaluations, test automation or review practices. Customer project delivery, including scoping and estimating integration work. Integration architecture and design patterns, such as enterprise integration patterns, API-led or layered API architecture, and event-driven integration. BPMN or other process-modelling notation. Building MCP servers, agent skills or agent workflows, and working with LLM APIs directly (for example Anthropic or OpenAI), including prompt and context design. Azure, especially integration and hosting services. What we offer: Real AI work in production: You build AI agent tooling that runs on real enterprise integration projects, not demos or prototypes. Work that reaches beyond one project: Your tooling is used by Frends consultants, partners and customers, so your impact scales well past your own projects. A hard, meaningful problem: Migrating from legacy integration platforms has always been slow, expensive, and risky. You get to change that with AI. Your integration expertise as an edge: Your integration