AI Engineer

About the opportunity

The Virtual Assistant team is on a mission to redefine how N26 customers interact with their bank. We believe the future of banking is conversational, and we are currently making a strategic investment in Generative AI to rethink our entire ecosystem. We aren’t just looking for someone to maintain chatbots; we want a Mid-Level AI Engineer who is ready to build the next generation of intelligent, multi-channel experiences.

As part of this team, you’ll help us bridge the gap between complex banking needs and seamless, human-like automation. If you’re passionate about moving beyond traditional flows and into the world of LLMs, RAG, and truly autonomous agents, we want to talk to you.

Your core mission is delivery, optimized for velocity and reliability: taking proven AI/ML models, LLM prototypes, and intelligent agents and integrating them directly into our product infrastructure. You will be obsessed with scalability, observability/evaluations, and speed, ensuring our AI solutions move swiftly and reliably from a proof-of-concept to a production-ready, highly monitored service that directly levels up the conversation experience of our customers.

This role requires a unique T-shaped skillset blending Data Engineering, Applied AI, and Backend engineering understanding.

• Drive Fast Time-to-Market (TTM):
  Work collaboratively with relevant stakeholders (Data Science, Machine Learning Engineering, etc.) to iterate quickly on prototypes, ensuring a seamless and rapid transition of validated models and LLM applications into the production environment.
• Productionize AI/LLM Prototypes:
  Act as the primary link between the DS/AI team's prototypes and the customer-facing product. You will refactor, deploy, and maintain robust, performant AI services (including LLM-powered features and intelligent agents) in a high-traffic production environment.
• Build Observability and Measurement Systems:
  Implement comprehensive observability solutions (evaluations, metrics, tracing, logging) for all deployed AI features. Design and execute a framework for measuring output quality, A/B testing different models/prompts, and iterating on performance in real-time.
• Architect for Reliability and Scale:
  Collaborate within the team to design and implement the technical architecture, data pipelines, and orchestration systems necessary to support AI services. Focus on building highly available, low-latency APIs and ensuring proper monitoring, logging, and failure handling across the entire lifecycle.
• Design and Deploy Custom AI Services:
  Collaborate with the Platform Engineering team on leveraging foundational models (like those from AWS Bedrock/Anthropic) managed by them. Your focus will be on designing, building, and deploying the surrounding application logic and custom components, such as Model Context Protocol (MCP) servers, Agents that interact with those foundational models.
• Apply Advanced AI Techniques:
  Directly implement and optimize advanced Applied AI techniques, such as Retrieval-Augmented Generation (RAG), Model Context Protocol (MCP), Agent-to-Agent (A2A) flows, and fine-tuning methodologies.
• Contribute to Discovery:
  Leverage your deep understanding of system constraints, data availability, and production feasibility to proactively identify and recommend new AI use cases, working closely with Product and Data Science.

• Proven experience building, deploying, and maintaining machine learning or LLM-based applications in a production environment.
• Hybrid Technical
  
  Skills:
  
  A strong blend of experience across the required disciplines:
  • Backend Engineering:
    Proficiency in developing scalable, reliable backend services (mainly Python, Kotlin) with a focus on API design and microservices architecture.
  • Applied AI Engineering & MLOps:
    Hands-on experience with LLM frameworks, vector databases, MLOps tooling, prompt engineering, and successfully implementing advanced techniques like RAG. Demonstrated ability to manage models and prompt updates safely.
  • Data Engineering:
    Familiarity with building robust data pipelines, managing feature stores, and ensuring data quality for…