About this role
Pinnipedia is a new Berlin startup building a cloud platform that automates and assists the creation of audit-ready IT-security concepts (e.g., BSI-Grundschutz, C5). We’re IGP-funded (2025/26) and co-develop with FU Berlin and pilot users from industry and security consulting.
We’re hiring an AI Engineer to turn messy inputs into structured knowledge and reliable answers. You’ll design and operate knowledge-graph + LLM (RAG) pipelines, model/ingest domain ontologies, and own evaluation so we can ship trustworthy features.
Tasks Knowledge graphs & data
Model the domain (ontology/taxonomy); build ETL into a graph store. Author queries (SPARQL/Cypher) and surface graph facts and relationships in features.
RAG & LLM integration
Design retrieval and answer generation workflows (indexing, chunking, reranking). Orchestrate prompts/tools; balance KG, vector search, and business rules.
Evaluation & quality
Define and track retrieval/answer metrics (e.g., precision/recall, faithfulness). Build test fixtures and regression checks; monitor drift and data quality.
Production & collaboration
Ship well-tested Python components (FastAPI jobs/services); document decisions; work from a clear backlog with PO and engineers.
Requirements Must-have
Strong Python and data engineering fundamentals. Hands-on with knowledge graphs (ontology design + queries) and a graph DB. Practical RAG experience (indexing, retrieval, evaluation). Testing mindset (pytest), version control, and clear documentation. English required (German nice-to-have).
Nice-to-have
Security/compliance awareness; prompt/agent tooling; spaCy/Transformers. Observability for ML/LLM systems; simple dashboards for quality metrics. Cloud basics (AWS/Azure), containers (Docker); CI/CD.
Benefits Hybrid, full-time with flexible scheduling; occasional on-site days in Berlin.
Competitive salary: 60.000–85.000 € base (more for exceptional senior profiles).
Small, focused team; direct collaboration with the Product Owner and Full-Stack Engineer.
Modern tooling, real ownership, and a learning budget for role-relevant training.
Impact: help SMEs meet rising security requirements with less friction.
Apply on JOIN with your CV (PDF) and a short note (max 200 words) describing how you would design a KG-backed RAG pipeline (ontology scope, indexing, retrieval, and evaluation you’d use).
Process: 20-min intro → 90-min practical (graph modeling + retrieval evaluation) → 45-min team chat → references. We review applications within 5 business days.