Senior AI Engineer

Your mission

About Us

We’re building a new kind of mortgage experience: transparent, digital, and truly customer-centric. We are a Berlin-based B2C Fintech founded by two repeat founders (YC alum, exit) with prior 10+ year careers in investment banking and strategy consulting. We recently raised investment to assemble a 30-person, exceptionally strong core team.

Our culture is built on ownership and clarity. We align with you on vision, set clear KPIs and resources together, and then trust your drive and expertise.

The Opportunity

Buying property (and financing it) is probably the biggest financial decision in someone’s life. It’s complicated, opaque and anxiety-laden.

As a Senior Backend Developer / AI Engineer, you will work alongside our CTO on building and evolving our AI-powered fintech product. The core architecture is in place, providing a solid foundation to build upon as we integrate advanced LLM capabilities, establish AI pipelines, implement intelligent automation, and ship production-ready AI features. You'll be instrumental in shaping how we leverage AI to transform financial services, with significant autonomy to design and implement cutting-edge solutions.

Your profile

What You’ll Do

What you will work on on daily basis

Designing and implementing LLM-powered features and intelligent automation workflowsBuilding robust AI pipelines with proper orchestration, error handling, and monitoringDeveloping RAG (Retrieval-Augmented Generation) systems for financial data processingCreating evaluation frameworks for LLM outputs and maintaining quality standardsIntegrating multiple AI models and managing model versioning/deploymentEnsuring AI safety, reliability, and compliance in financial applicationsBuilding evaluation frameworks for AI outputs -- measuring extraction accuracy, classification precision, and hallucination rates per document typeSetting up MLOps practices: model versioning, automated quality gates before model swaps, experiment tracking (Weights & Biases, MLflow, or similar)Establishing data flywheel patterns -- using production results and human corrections to continuously improve extraction qualityOptimising for the cost/accuracy/latency triangle -- choosing the right model and approach for each pipeline stage

What You Bring

Hard Requirements

LLM Framework Expertise: Deep experience with LangChain/LangGraph (JS preferred, Python acceptable). You've built multi-step agent workflows, not just chatbots. You understand state management, structured output extraction, and multi-agent orchestration. Production Engineering: Expert-level TypeScript/Node.js preferred. Strong backend experience in Go, Java, C#, or Kotlin also accepted -- you'll ramp on TS quickly. Python-only backgrounds require demonstrated production backend experience (not notebooks/scripts). You've built, deployed, and operated services in a regulated or high-stakes environment.Workflow Orchestration: Production experience with Temporal.io (TypeScript SDK) or similar durable workflow engines. Reliable, long-running pipelines with retries, error handling, and state management.RAG, Fine-tuning & Model Selection: Experience with RAG architectures (vector databases, embedding strategies, hybrid retrieval) AND practical understanding of when to fine-tune vs. few-shot prompt vs. RAG. You have a decision framework for choosing the right approach -- not just defaulting to one tool for everything. Hands-on fine-tuning experience preferred (Hugging Face, LoRA/QLoRA, or cloud fine-tuning APIs).MLOps & Experiment Tracking: Experience with model versioning, automated quality gates, and structured experiment tracking (Weights & Biases, MLflow, or similar). You don't ship a model swap without measuring its impact. You've set up or contributed to pipelines that track model performance over time and prevent regressions.Model Evaluation & Quality: You've built evaluation frameworks for AI outputs -- measured precision/recall, confidence calibration, and automated quality monitoring. Not "it looks right" but systematic measurement.

Additional Technical Skills

Experience with prompt engineering and prompt management systemsKnowledge of AI observability tools (LangSmith, Helicone, or custom solutions)Understanding of AI safety, hallucination detection, and output validation techniquesDocument Processing & Intelligence: Experience with document extraction pipelines -- OCR, classification, structured data extraction from PDFs/images. Experience with cloud document AI services (GCP Document AI, AWS Textract, Azure DI). You know the difference between throwing an LLM at a document and building a robust extraction system.Multimodal AI: Experience with vision-language models for document understanding (GPT-4V, Gemini Vision, Claude Vision) -- structured extraction from complex layouts, not just image captioningCost Optimization: Model routing, caching, batching, right-sizing models per task

Mindset & Approach

AI-Native Thinking: You think in terms of probabilistic systems, understand LLM limitations, and know how to design robust systems around non-deterministic outputs.Pragmatic Innovation: You know when NOT to use an LLM. You can reach for a fine-tuned small model, a rules engine, or a cloud AI service when it's the better tool.Ownership Mentality: You're proactive, resourceful, and accountable for your work—you don't wait to be told what to do.Team Player: You collaborate well with others, can explain AI concepts to non-technical stakeholders, and elevate those around you.Quality-Focused: You understand the importance of evaluation, testing, and monitoring in AI systems and build with observability in mind.Continuous Learner: You stay current with the rapidly evolving AI landscape, experiment with new models and techniques, and are energized by pushing the boundaries of what's possible.Engineering Rigor in the Age of AI Tools: Coding agents can write code but can't own production reliability. You care about error handling, type safety, testing, and observability -- even when the industry says "just ship it."AI Co-ownership: You don't just execute tasks -- you evaluate approaches, challenge assumptions, and come back with "I tested three options, here's my recommendation." You co-own AI strategy, not just implementation.

Why us?

What We Offer

Competitive salaryParticipation in our employee options programmeOn-site in Berlin, with flexible remote work as you see fitHigh autonomy, low politics, fast decision-makingThe chance to help build something new, with real influence on how people buy homes in Germany!

About Us

We are an equal opportunity employer. We value diversity and are committed to an inclusive environment for all employees. If you are excited about the role but don't meet 100% of the requirements, we would still love to hear from you.