Senior AI / Knowledge Graph Engineer (m/f/d)

Pinnipedia, a Berlin startup, is looking for a Senior AI/Knowledge Graph Engineer to build a cloud platform that automates the creation of audit-ready IT-security concepts. The role involves owning the end-to-end pipeline that turns unstructured documents into a validated, queryable knowledge graph using LLM extraction agents, graph databases, and evaluation frameworks.

• Build LLM extraction pipelines for document chunking, property/relationship extraction, cross-chunk reconciliation, and gap detection.
• Design knowledge graph schema as typed Pydantic models, including Cypher access patterns and indexing strategy, and manage schema evolution.
• Develop deterministic rule engines for cases where code outperforms LLM judgment, maintaining clear contracts between deterministic and probabilistic components.
• Ensure data validation and quality through schema enforcement, required-property contracts, audit trails, and eval harnesses (including LLM-as-judge).
• Handle live data operations: backfills, coordinated migrations across relational and graph stores, observability on extraction throughput/quality, and incident response.

• 5+ years shipping data/AI systems to production with real customers, including on-call experience for live pipelines.
• Strong Python (typed, modern) and SQL, with comfort using PostgreSQL under load.