Process Automation Engineer

We’re a global team of over 400 people, working together to push the boundaries of open-source technology and multi-cloud solutions. Our vision is to help developers, builders, and creators bring their ideas to life with speed and simplicity, by providing a cloud data platform that makes open-source databases, search, streaming, and application infrastructure easily accessible to everyone. The Role We're building one shared internal automation platform, and hiring the engineer to build it. Aiven runs on a lot of processes nobody wrote down. The worst of it sits between systems: an approval living in three systems where one would be enough, numbers copied by hand from one place to the next. Your role is to work out what's a simple integration, what actually needs AI's judgment, and how the two run together. Your Mission: Build the Automation Platform, Starting in Finance You will own the automation platform, starting in Finance, the function with the least room for error and the most to gain from getting this right. Two other people in Enterprise IT are already automating parts of Finance. You'll give that work a shared platform and reusable capabilities, so it scales past one-off workflows. Prove it there, and it can grow into a company automation platform. Here's the part most automation roles don't offer: we have a big say in picking the problems. The role is in the Data Platform & Enterprise IT team. We make the whole company observable through data and own a lot of the systems it runs on. We are in a perfect place to figure out what needs automating in the business and actually do something about it. And people want us to help them: we have strong relationships across Aiven and more demand than we can serve, so what you ship gets adopted. This is a senior individual-contributor role for someone who treats automation as infrastructure. AI is central to how you'll work: you'll build with agents, automation platforms, and MCPs every day. You always reach for the simplest solution, not the trendiest. If certain systems require tons of customisation, we'll push for changing the system, not build around it. This isn't legacy RPA, recording desktop clicks or automating individual tasks. The platform might use visual tools, but everything that matters is production software: versioned, tested, observable, owned. In concrete terms, this means you'll: - Own the automation platform: Build, run, and scale our orchestration platform as real infrastructure, production workflows with observability, error handling, and retries, not demos that quietly rot. We run n8n today; you own where it goes. - Turn it into a platform, not a services queue: Find the primitives Finance needs and ship them as reusable capabilities, so you're not reinventing reconciliation and the next workflow is a paved road, not a six-month project. - Start where getting it wrong isn't an option: Finance. Today, closing the books or reconciling bank transactions still means moving data by hand between systems like NetSuite, Zip, Atlar, and Rydoo. You take workflows like procure-to-pay and the month-end close and rebuild them so they run end to end, with approvals and exception handling where financial controls or human judgment need them. If it survives Finance's deadlines and its auditors, it scales anywhere. - Put agents to work where they earn it: Design human-in-the-loop and autonomous agent workflows for the cases where an agent genuinely beats a deterministic flow. - Partner across the systems you orchestrate: Work alongside the engineers and specialists who own the underlying platforms (NetSuite, billing, payments); you orchestrate the flows across them, they own what's underneath. What good looks like: Within your first six months, at least one manual, deadline-driven Finance workflow (the kind that today runs through a spreadsheet, a Slack ping, and someone's memory of the exception cases) runs on your platform instead: observable and auditable, routine failures recover safely, and the exceptions reach the right person without anyone watching the process by hand. What We're Looking For Must-have: - A builder with high agency: You've seen a broken process nobody told you to fix, and fixed it. You start, you decide, you ship. You don't wait for permission or a perfect spec. - A systems thinker: It's easy for you to break a process down into its parts, put it in a diagram, and think about the system as a whole. - You can find out how work really happens: You sit with the people doing the job, tell the real process apart from the official one, map the normal path and the exceptions, and turn something nobody wrote down into something you can build, without losing the detail that makes it work. - A fast learner, across domains: You can go from knowing nothing about procurement or the month-end close to automating it, because you get genuinely curious about how things work. - Built and run real automation in production: n8n, Temporal, Airflow, Prefect, or you wired systems together with your own code and APIs and owned the result. You think in API contracts, webhooks, idempotency, retries, secrets. Not a POC you ran once, something people depended on. - Real software discipline: Version control, tests, not scripts that rot. - AI is already how you build: Not something you need convincing about. You use Claude or similar, MCP, and agent orchestration to ship real things. You've built workflows where models read information, make decisions, or take action, and you know how they fail. You reach for an agent when it's right and a plain workflow when it isn't. - Fix the process, not the symptom: When someone asks you to automate everything from the ticket onward, your first instinct is to ask why there's a ticket at all, then redesign the flow end to end. And when something breaks, you don't stop at the fix. You dig into why it happened and change that too. Strong advantages, not requirements: Finance familiarity: Familiarity with finance systems (NetSuite, billing, payments) at the integration layer helps, and so does knowing the finance domain itself — the month-end close, reconciling bank transactions — not just the systems. We'd still rather hire an exceptional automation engineer who can learn the finance domain than a finance-systems specialist who can't build the platform. Finance is where you start.