AI Engineer: Build agentic AI on Google Cloud that customers run in production

Gapps; Helsinki; Hybrid; EUR 56,000 - 73,000; Full time

Job summary

Gapps Group is seeking an experienced AI Solutions Engineer to design and build agentic workflows and LLM applications on Google Cloud for real customer deployments. You will work closely with customers from problem definition through production, contributing to data pipelines and integration efforts.

Based in Helsinki with hybrid work, you will collaborate with the AI Solutions team and leaders to deliver production-grade AI capabilities, mentor new engineers, and drive architectural decisions

Qualifications

• Hands-on Google Cloud experience.
• Experience delivering AI engineering in production.
• Ability to design and build agentic workflows.

Responsibilities

• Design and build AI solutions on Google Cloud for customers.
• Build LLM applications, RAG solutions and integrations with existing systems.
• Own technical workstreams and collaborate with cloud architects on security and production readiness.
• Explain technical choices to customers and leadership.
• Help grow and mentor the AI Solutions team.

Skills

Google Cloud experience

AI engineering

Client interaction

Architectural design

Tools

GCP

Vertex AI

Job description

Helsinki (hybrid) · Full-time, permanent · Apply by 29.11.2026

You've built an agent that works. Then it sat in a sandbox for months, waiting for a budget and for someone willing to own it. At Gapps, you’re the one who gets it out of the sandbox. You work out the real problem with the customer and build the solution on Google Cloud yourself. Then you watch it go live in their daily operations.

The role at a glance

What you'll do: Design and build AI solutions for Gapps' customers on Google Cloud: agentic workflows, LLM applications, RAG, and the integrations that connect them to real systems. You talk with the customer about the problem first, then build the solution yourself.

Pay: Base salary 5,000–6,500 € a month, depending on your experience. On top of that, a shared team billing bonus paid monthly: 900 € a month at target, uncapped, typically 10,000–20,000 € a year.

Where: Hybrid, based at our Helsinki HQ in Kalasatama. Some customer work happens on-site.

Team: You'll work closely with Elmo Visuri (Tech Lead) and report to Jani Muuronen (AI Business Unit Lead).

Growth: AI Solutions is a young team, and you'd be among its first engineers. You help decide how Gapps builds AI, and as the team grows you can take on more technical ownership or mentor the people who join after you.

This role probably isn’t for you if you’d rather stay cloud-neutral, or you want to code from a ticket queue without meeting customers, or draw architectures for someone else to build. Google Cloud experience matters a lot here. If your experience is mainly on another cloud, show us how quickly you’ve picked up new platforms before. And if your AI work so far is mostly demos and side projects, this will be a big jump.

Why Gapps is your next move

Maybe you're an AI or ML engineer in a large organization. You've built things that work, but every step toward production needs one more approval or one more budget round.

Or maybe you're at a consultancy that works across every cloud. You rebuild the same proof of concept for one customer after another and rarely see one reach the people it was built for.

At Gapps, the work is focused on one ecosystem, and the demand is already there. Our AI Adoption team trains customers' people and spots where AI could help. When a customer decides they need an agent, someone has to build it. That's you.

Elmo Visuri, our Tech Lead, on how his projects usually go:

"The way it works for me is I talk to someone higher up and sell them the good idea. Then I make them a design for it: there's your architecture, the one you want. And then I'm like, hey, I can build it."

What the role involves

You work with customers from the first conversation about the problem until the solution runs in their environment, and you keep improving it after that. Some projects are longer engagements where you're part of the customer's extended team. Others are short: a proof of concept, a specific build, or a quick prototype of one stakeholder's idea.

Discovery and solution design

• Understand the customer's business and find where AI creates measurable value
• Turn business needs into technical requirements and a plan that can actually be built
• Push back when the business value or the risk doesn't justify building a use case
• Support presales with feasibility checks and solution ideas

Building and delivery

• Design, build, test and deploy AI applications and agentic workflows, including multi-agent solutions
• Build LLM applications, RAG solutions and integrations with existing systems, mostly on Google Cloud
• Contribute to ML components and data pipelines when a project needs them
• Own technical workstreams and work with our cloud architects to keep solutions secure and production-ready

Customers and team

• Explain technical choices and limits to everyone from operational staff to senior leadership
• Help customers build their own understanding, so they rely on you less over time
• Shape how Gapps delivers AI solutions and help new engineers get going as the team grows

A typical week

We asked Elmo what his week looks like. His answer: "All of this, mixed, at random weights."

One day it's an agent platform for a customer service process. The next it's data pipelines, or sparring a customer's team on GCP. Sometimes you're invited to a meeting about MCP connectors with no agenda at all, and the job is to show up and be useful. A senior leader might ask for a hands-on session to build their own workflows with you. Your mix will depend on the customers and projects you land in.

Who you'll work with

• Jani Muuronen, AI Business Unit Lead and your manager: business priorities and where AI Solutions is heading
• Elmo Visuri, Tech Lead: technical direction, architecture and day-to-day building
• The rest of the AI team: Lauri Raitasuo (Lead Solutions Architect), Leena Peltomaa (Senior AI Solutions Consultant), Sini Liu (Senior AI Change Lead) and Atte Front (Project & Change Manager). Meet them on our blog.
• Our Google Cloud team: architecture, security and the customer's cloud environment
• The customer's people, from technical specialists to business leadership

Your first year

After 30 days you know our AI Solutions offering, current projects and the team. You've met the customers you'll work with, joined presales discussions, and started contributing to solution planning and building.

After 90 days you're working in one or more customer projects, moving clearly defined technical work forward on your own, and you've shipped at least one component into a customer's environment.

After 12 months you've delivered meaningfull work across several customer projects and work independently on AI engineering assignments. You're helping build our AI engineering practice, possibly recruiting and onboarding new engineers, and maybe you've built something that improves Gapps' own operations too.

About Gapps

Gapps is a Finnish cloud and digital transformation consultancy. We've worked with Google since 2010, grown into a multi-million-euro company, and been named Google Cloud Country Partner of the Year for Finland. Our team works across cloud infrastructure, AI, managed services, digital workplaces and security.

We're shifting from one-off projects and license revenue towards productized services, long-term partnerships and AI solutions. In 2026, Vesa Sironen returned as CEO with the goal of making us Finland's most trusted partner for AI and cloud services. AI Solutions is where much of that shift is happening.

We're small enough that your work is visible and your decisions matter. The culture runs on high professional standards, low hierarchy and responsibility given early. As Jyrki Rautio, who leads a large migration project, puts it:

"You get responsibility. You have to be able to carry it. But you are not alone."

Salary and practicals

Compensation. Base salary is 5,000–6,500 € a month. Where you land depends on your relevant experience and skills, and we're happy to talk it through. You also share in a team billing bonus, paid monthly: 900 € a month at target billing, uncapped, and typically 10,000–20,000 € a year. The model may be updated for 2027; we’ll walk you through the exact mechanics and numbers during the process.

This is a permanent, full-time position, starting preferably in January 2027.

Location. Hybrid, based at our Helsinki HQ in Kalasatama. You need to be based in Finland and already have the right to work here. The office has ergonomic workstations, quiet focus rooms, a gym and snacks, and we equip a proper home workstation for your remote days.

Health and wellbeing. Extensive occupational healthcare and insurance covering both physical and mental wellbeing, plus lunch, sport and culture benefits.

Professional growth. Google Cloud certifications completed on company time, support for other relevant development, and dedicated time built into your working hours for studying and developing your expertise.

Language. Fluent English. Finnish is a big plus, since many customer conversations happen in Finnish.

What matters most

Here's an honest look at what matters most for this role. You don't need to check every box, but these make the biggest difference.

• Hands-on Google Cloud experience. You've built and deployed things on GCP and know what it takes to configure them inside a customer's existing environment. Certifications help, but what you've built counts more.
• AI solutions in production. You've delivered AI that people use in real operations, beyond experiments and demos, and you know what security and maintenance look like after go‑live.
• Hands‑on building. Agents, agentic workflows, LLM applications, RAG, API integrations: you can design, build and troubleshoot them yourself. AI coding tools are fine, as long as you understand what they produce.
• Business sense. You look for the real problem behind a technical request, and you say so when a use case isn't worth building.
• Credibility with customers. You can explain technical options and limits to people outside tech, and an unplanned meeting doesn't throw you.

Nice to have: Gemini Enterprise deployments, Vertex AI, MCP, multi-agent orchestration, ML engineering and data pipelines, and AI security and governance in production.