Senior Applied AI Engineer
N-iX · Spain
You apply on the site where the job is posted. I never handle applications.
European Union
Work type:
Office/Remote
Technical Level:
Senior
Job Category:
Software Development
Our client is a fast-growing European fintech company in the business spend management space. We're looking for a Senior Applied AI Engineer to design, build and ship customer-facing AI features end-to-end.
Responsibilities
Design, build and ship customer-facing AI features end-to-end, including RAG system design (chunking, embedding selection, retrieval, re-ranking) and agentic workflows.
Ship external customer-facing LLM-based features into production, with safe production deployment practices.
Do hands-on evaluation-pipeline and observability work: drift detection, quality monitoring.
Reason independently about retrieval architecture using strong data fluency.
Apply real engineering rigor to the Context Development Lifecycle (CDLC): generate, evaluate, distribute and observe the context that powers AI agents.
Multiply the output and quality of the squad you join, and share patterns/practice beyond your immediate team so adoption compounds across the organisation.
Embed directly into a client squad as a hands-on Individual Contributor.
Requirements:
Python (Python-first).
RAG system design: chunking, embedding selection, retrieval, re-ranking.
Agentic workflows; safe production deployment.
Proven experience shipping external customer-facing LLM-based features to production (not prototypes).
Hands-on evaluation-pipeline and observability work: drift detection, quality monitoring.
Data fluency to reason about retrieval architecture independently.
Agent & harness engineering fluency — designing guardrails, context and verification layers for safe, reliable AI-assisted delivery; concrete evidence of having built/operated this kind of tooling, not just used it.
Comfortable with the client's current daily tools: Claude Code and GitHub Copilot.
Nice-to-Have:
Specific tooling such as AWS Bedrock, OpenAI APIs, or Langfuse (or comparable eval platforms).