Lead AI Engineer (Remote)
Software Engineering, Data Science
Berlin, Germany · Barcelona, Spain · Romania · Hungary · Estonia · Portugal · Bulgaria · Greece · Slovenia · Latvia · Cyprus · Warsaw, Poland · Belgrade, Serbia · United Kingdom
We are looking for a Lead AI Engineer to own the technical direction of AI systems across Finom and lead the engineers who build them.
This is a leadership role from day one. You will set the technical and quality bar, lead a team of AI engineers, and stay hands-on in the code. You have done this before — we are not looking for someone stepping into technical leadership for the first time.
You will build AI services for Finom customers, and internal platform solutions that allow other teams to ship AI capabilities of their own. You take architecture responsibility for the services you and your team develop, along with the alignment that comes with it — translating between business intent and technical reality, and challenging product decisions when they are wrong.
This is not a research role. It is a hands-on engineering leadership role focused on production-grade AI capabilities that create clear value for customers and the business.
What You Will Be Doing
- Lead a team of AI engineers — set direction, review designs, and grow their technical scope through review, pairing, and design guidance
- Own the architecture for the services you and your team develop — and write down the decisions, tradeoffs, and rejected alternatives so they can be reviewed and challenged
- Build and ship AI-powered customer and internal solutions using LLMs, RAG, tool calling, workflows, and agentic patterns
- Own AI systems end to end — problem framing, implementation, evaluation, deployment, monitoring, and iteration
- Develop scalable and reliable inference pipelines with strong attention to latency, cost, security, and observability
- Set the evaluation and quality bar — offline evals, online signals, failure analysis, and continuous improvement loops that other teams adopt rather than renegotiate per project
- Design AI systems with auditability, model governance, and EU AI Act obligations built in from the start rather than retrofitted
- Drive AI platform and tooling decisions that improve reuse, speed, and consistency across teams
- Partner with solution managers, domain teams, and engineers to integrate AI into real workflows rather than isolated demos
- Tell a product stakeholder when the thing they asked for is the wrong solution, and be persuasive about the right one
- Negotiate scope and sequencing with domain teams that have their own roadmaps and no obligation to yours
- Decide what to stop doing — deprecate, simplify, or kill approaches that the evidence no longer supports
- Shape the roadmap rather than only execute tickets
Who You Are
- A leader who earns authority through technical credibility rather than title
- Still an engineer at heart — close enough to the code and the designs to have an opinion worth defending
- Comfortable being accountable for outcomes you did not personally build
- Someone who has held a technical position against pushback, revised it when the pushback was right, and can tell the difference
- Direct with stakeholders and with your team — you surface problems early, including the ones that reflect badly on your own decisions
- Decisive under ambiguity: you make the call on incomplete information and revisit it when better information arrives
- Clear in writing — able to make a technical argument that a non-engineer can follow and act on
- Product-minded and focused on real user outcomes, not just model outputs
- Curious, low-ego, and biased toward action; motivated by what the team ships, not only by what you ship yourself
Must-Haves
- Proven experience leading a team of engineers, formally or as a tech lead, with responsibility for what the team delivered
- Track record of owning the technical direction of a system used by teams other than your own
- Experience changing a product or business decision through technical argument rather than escalation
- Experience working directly with non-technical stakeholders on commitments and tradeoffs, not through a manager as intermediary
- Experience growing other engineers through review, mentoring, or design guidance
- Strong ownership mindset and the ability to create clarity in genuine ambiguity
- Proven experience building and deploying AI systems in production
- Strong Python and software engineering fundamentals
- Hands-on experience with LLM applications, including some of: RAG, tool use, agents, prompt engineering, evals, structured outputs, guardrails, or fine-tuning
- Experience integrating AI systems into backend or product workflows
- Ability to design meaningful evaluation, monitoring, and continuous improvement loops
- Experience with cloud infrastructure and containerized deployments
- Strong grasp of the fast-moving AI landscape, with the ability to turn relevant advances into practical product and engineering decisions
- Fluent English (C1)
Nice-to-Haves
- Experience in fintech, financial services, risk, compliance, or operations-heavy environments
- Experience with applied ML beyond LLMs, such as classification, anomaly detection, ranking, or document intelligence
- Experience with vector databases, knowledge systems, and retrieval infrastructure
- Experience with model benchmarking, experimentation frameworks, and cost or latency optimization at scale
- Background in startups or as a founder
- Contributions to open-source or visible side projects in AI
Example Tech Stack
- Languages: Python, SQL, noSQL
- LLM / AI: OpenAI, Anthropic, LangGraph, Hugging Face, Ollama, PyTorch, OpenClaw
- Patterns: RAG, tool calling, agent workflows, eval pipelines
- Infrastructure: Docker, Kubernetes, AWS / GCP / Azure
- Data / Platform: Vector databases, event-driven systems, APIs, observability tooling
You do not need experience with every item, but this role will likely involve technologies such as:
Tech Stack
- Languages: Python, SQL, noSQL, .NET (optional)
- LLM / AI: OpenAI, Anthropic, LangGraph, Hugging Face, Ollama, PyTorch, OpenClaw
- Patterns: RAG, tool calling, agent workflows, eval pipelines
- Infrastructure: Docker, Kubernetes, AWS / GCP / Azure
- Data / Platform: Vector databases, event-driven systems, APIs, observability tooling
What Success Looks Like
In your first 3 months
- You have enough context on the team and the business domains it serves to know which problems are worth solving
- You have taken architectural ownership of your team's services, including the decisions you inherited
- The team's priorities are clear and defensible, and you are who product and business stakeholders come to for them
- You have identified the biggest constraint on the team's delivery — technical debt, unclear ownership, missing skills, or process — and started removing it
By 6 to 12 months
- Your team has delivered significant AI capabilities to Finom customers, with measurable impact through revenue uplift, cost savings, productivity gains, or risk reduction
- The team delivers predictably: commitments you make on its behalf hold, and when they will not, stakeholders hear it from you early
- The architecture and standards you set are documented, adopted, and used by teams beyond your own
- Quality is systemic rather than heroic — failures are caught by evals and monitoring rather than by customers, and the team debugs its own production issues
- Engineers on your team have grown measurably in scope, and several can lead significant workstreams without you in them
- You shape the AI roadmap together with product and business leadership rather than receiving it, and your "not this way" or "not yet" carries weight