Open position · Full-time · Helsinki
Forward Deployed Engineer
Close the gap between the business idea and the working system.
The role
Most AI projects fail in the gap between the business idea and the working system. Your job is to close that gap. You’ll sit with business users to understand how work actually happens, decide with them what’s worth changing, and then build, test and improve the solution until it holds up in everyday use.
This means you’ll split your time between understanding and building. One day you’re mapping a finance process with its owners; the next you’re connecting systems, building an agent and measuring how well it performs. You’ll help clients understand what’s feasible, what needs to change in the way they work, and what it takes to move from a promising prototype to something people rely on.
You’ll work inside client teams, close to the people who use what you build. What you build sets the pace for the whole programme, and what you learn shapes where it goes next.
What you’ll work on
- Understand the work in detail: map how a process runs task by task, including the systems, data and decisions involved, together with the people who do the work.
- Shape the solution: turn the priorities agreed with leadership into concrete solutions, defining what to build, for whom and how success is measured.
- Build and test: create prototypes, integrations, AI workflows and agents on the client’s own AI platforms, such as Microsoft Copilot and Claude, working with real client systems and data.
- Make the new process work: build and automate the redesigned process with our AI strategists, including where human judgement stays in the loop.
- Run it in production: deploy solutions, monitor quality, reliability and cost, and handle security and data access with the client’s IT.
- Prove the value: measure real performance and business impact, explain trade-offs clearly to decision-makers, and give our strategists the evidence for scaling.
- Scale and reuse: work with Witted’s engineers to turn proven solutions into robust systems, and turn what you learn into components and patterns for the next client.
What you bring
You have a solid foundation in software or data engineering and practical experience building with AI: LLMs, agents, retrieval, workflows. You can take an unclear business problem, ask useful questions and turn it into something people can try.
You’ve built with more than one AI platform, for example Microsoft Copilot Studio and Azure OpenAI, or Claude. You know their strengths, limits and costs, and you choose what fits the client’s environment and the task, not what you happen to prefer.
You’re comfortable with APIs, integrations, cloud environments and the messy reality of enterprise data. You know how to test AI systems, and you recognise where they need safeguards, monitoring or human judgement.
Just as importantly, you’re curious about how businesses work. You can talk about costs, productivity and decision-making as comfortably as about architecture. You can explain technical choices clearly to senior leaders, listen to the people doing the work, and change course as you learn.
It’s a plus if you have
- Experience from a client-facing role such as consulting, solutions engineering or product work.
- Experience of building AI solutions for functions like finance, operations, customer service or software development.
- Experience of measuring the cost and performance of AI systems, including model and token costs.
- Fluent Finnish, plus English. Norwegian or Swedish is a bonus.
What we offer
- The chance to shape a new company from the start, working closely with its four founding partners.
- Work where you both decide what to build and build it yourself.
- Close collaboration with Witted’s engineers and data specialists.
- Clients among the largest companies in Finland and Norway.
- Room to develop our technical methods, assets and tools as we grow.
Interested?
Send us your CV or LinkedIn profile and a short example of something you’ve built. Tell us what problem it solved, what you learned and what you would do differently.
info@