AI Solutions Technical Engineering Manager

FDJ UNITED; Stockholms kommun; On-site; SEK 1,600,000 - 2,600,000; Full time

Job summary

FDJ UNITED is seeking a senior AI Solutions Technical Engineering Manager to lead the delivery of AI- and data-enabled products and platforms from shaping to production outcomes. You will steer engineering leadership, AI delivery, and stakeholder alignment while turning ambiguous ideas into actionable plans.

You will oversee cross-domain teams (AI/ML, data, software, platform) to deliver trusted, production-grade AI capabilities, balancing hands-on work with strategic decisions under tight

Qualifications

• Proven experience leading engineering delivery for data-heavy AI products.
• Strong understanding of ML, LLMs, data dependencies, evaluation.
• Comfortable operating in federated environments with shared ownership.

Responsibilities

• Lead multidisciplinary teams (AI/ML, data, software, platform) to deliver AI capabilities.
• Split time between leading the team and hands-on design reviews.
• Define architecture, evaluation, and deployment strategies for LLM systems.
• Own delivery end-to-end of AI workloads and products.
• Represent AI delivery to senior stakeholders.

Skills

AI/ML systems

LLM systems

Tool use

MCP gateway

Kubernetes

AWS

CI/CD

Observability

Stakeholder alignment

Communication

Tools

Confluence

Jira

Teams

ClickHouse

Bedrock

OpenRouter

Job description

At FDJ UNITED, we don't just follow the game, we reinvent it.

FDJ UNITED is one of Europe’s leading betting and gaming operators, with a vast portfolio of iconic brands and a reputation for technological excellence. With more than 5,000 employees and a presence in around fifteen regulated markets, the Group offers a diversified, responsible range of games, both under exclusive rights and open to competition. We set new standards, proving that entertainment and safety can go hand in hand. Here, you’ll work alongside a team of passionate individuals dedicated to delivering the best and safest entertaining experiences for our customers every day.

We’re looking for bold people who are eager to succeed and ready to level-up the game. If you thrive on innovation, embrace challenges, and want to make a real impact at all levels, FDJ UNITED is your playing field.

Join us in shaping the future of gaming. Are you ready to LEVEL-UP THE GAME?

We’re looking for an AI Solutions Technical Engineering Manager to lead the delivery of AI- and data-enabled products and platforms from early shaping through to measurable outcomes in production. You’ll operate at the intersection of engineering leadership, AI delivery, and stakeholder alignment, turning ambiguous ideas into structured plans, unblocking teams, and ensuring solutions are operationally viable.

You will work across domain teams in a federated model, bringing clarity around ownership, interfaces, and accountability. This is a senior role for someone who can drive innovation while delivering under tight timelines and high expectations.

The role

You will lead multidisciplinary engineering teams (AI/ML, data, software, platform) to deliver AI capabilities that are trusted, adopted, and production-grade.

Day to day, you're splitting time between running the team and doing real technical work alongside them — reviewing designs and code, prototyping when it's faster than explaining, and making the calls on deployment, routing, evaluation, and monitoring for our LLM and ML systems. That includes when to use semantic routing versus static routing, when OpenRouter‑style multi‑provider setups make sense, and how Claude and other models stay integrated cleanly across different tech stacks rather than bolted on inconsistently.

You’ll own technical direction of our text-to-SQL platform. You're also the technical backstop for agent and MCP work: tool‑use design, orchestration patterns, and knowing when MCP is the right call versus overkill.

On models, you decide whether a problem needs fine‑tuning or whether better prompting, retrieval, or routing gets there faster and cheaper; backed by real evaluation data, not intuition.

Beyond that, you keep what we ship production‑grade — secure, observable, cost‑sensible, reliable. You manage the people side properly too: coaching, growth, honest feedback. And you represent AI delivery to senior stakeholders, staying calm and credible under tight timelines and open questions.

What you’ll be doing

• Someone who has actually built and shipped AI/ML systems themselves, recently enough that they can still read and write code and hold their own in a design review
• Real depth on LLM systems — RAG, agentic architectures, tool use, evaluation, guardrails, model lifecycle — not just familiarity with the terms
• Hands‑on experience with multi‑model or multi‑provider setups — routing (semantic routers, OpenRouter or equivalents), fallback strategies, and cost/latency trade‑offs across providers
• You're also the technical backstop for our KAIT platform - our LibreChat‑based internal AI tool, covering MCP gateway architecture, agent orchestration, and integrations across Confluence, Jira, Teams, and ClickHouse.
• Practical experience with MCP‑based integrations and agent orchestration, including the judgment to know when an agentic approach is overkill
• A working understanding of fine‑tuning versus prompt/retrieval‑based approaches, and the evaluation rigor to justify the choice with data
• Experience leading a small senior team where you were still the most technical person in the room, or close to it
• Comfortable making architecture decisions without a clear playbook, and confident enough to defend them on technical merit
• Solid production engineering fundamentals — AWS, Kubernetes, CI/CD, observability — enough to have an opinion, not necessarily to build it yourself day to day
• Genuinely plugged into what's happening in AI right now, and able to tell the difference between a real capability shift and a marketing claim
• Communicates well enough to make technical trade‑offs land with non‑technical senior stakeholders, without dumbing anything down
• Own delivery of AI workloads and AI‑enabled products end‑to‑end: from discovery through build, launch, and iteration
• Convert loosely defined ideas into delivery plans, milestones, and measurable outcomes

What success looks like

• AI ideas actually make it to production — not stuck in prototype limbo — and you can point to real business impact, not just "it's live"
• The team ships fast without cutting corners on evaluation, security, or cost — because you're close enough to the work to catch problems before they become production incidents
• Your engineers see you as a technical peer they'd elevate a hard problem to, not just someone tracking their tickets
• Text‑2‑SQL, the agent/MCP work, and model integrations stay reliable and well‑governed as they scale — not held together by one person's tribal knowledge
• When trade‑offs get hard — speed vs. quality, build vs. buy, fine‑tune vs. prompt — you make the call, back it with data, and stakeholders trust it even when the timeline is brutal

Experience and capabilities (essential)

• Proven experience leading engineering delivery for data‑heavy, AI‑enabled, or platform products
• Strong understanding of modern AI concepts: ML systems, LLMs, data dependencies, evaluation, and operational risks
• Track record managing complex deliveries across multiple teams and stakeholders
• Comfortable operating in federated/domain‑oriented environments with shared ownership
• Excellent communication: able to align senior stakeholders and guide teams through ambiguity
• Solid grasp of production engineering fundamentals: cloud, reliability, security, monitoring, CI/CD

Technical environment

(Not hands‑on coding daily, but technically credible)

• Observability: metrics, logs, tracing; dashboards and alerting disciplines
• Delivery experience in regulated or high‑stakes environments (financial services, gambling/gaming compliance)
• Working knowledge of AI governance and where regulation like the EU AI Act is heading
• Direct experience with Bedrock or similar enterprise LLM platforms, including cost and access controls across multiple teams and tenants
• Experience with access‑control‑sensitive text‑to‑SQL or similar natural‑language‑to‑structured‑query systems
• Some visible engagement with the AI community — talks, writing, open source — as evidence you're tracking the field, not just reading about it secondhand