AI Knowledge Engineer

 AI Knowledge Engineer

The Role

We are hiring an engineer who lives at the intersection of knowledge graphs, LLMs, and shipping software fast. You will be building systems that map and reason over large codebases, integrating with cloud infrastructure, and turning architectural understanding into working product; not writing design docs that gather dust.

This is a hands-on delivery role. You will be using Claude Code as your primary development tool and expected to move at the pace that enables.

What You'll Be Doing

• Designing and building knowledge graph pipelines: ingestion, schema design, traversal, and query optimisation (Neo4j / property graphs)

• Working with embeddings and LLM APIs to enrich graph-based reasoning: you understand when to embed, when to traverse, and why RAG alone doesn't cut it at scale

• Integrating with real-world infrastructure: pulling context from AWS, GCP, Azure, and Kubernetes clusters,  not operating them day-to-day, but understanding their architectures well enough to extract meaningful knowledge from them

• Working with IaC artifacts (Terraform, Pulumi, CloudFormation) as data sources, parsing, interpreting, and mapping infrastructure-as-code into structured representations

• Delivering working software in short cycles using Claude Code as your core development workflow

What We're Looking For

Must-haves:

• Deep, practical experience with knowledge graphs: you have built them, not just read about them. You can talk fluently about ontology design, graph traversal strategies, and when a graph model beats a relational or document model

• Strong understanding of embeddings - vector spaces, similarity search, chunking strategies, and the trade-offs between embedding-based retrieval and structured graph queries

• Solid working knowledge of LLMs: prompt engineering, context window management, tool use, and how to build reliable systems on top of non-deterministic models

• Proven ability to ship software quickly ; we don't care about your CS degree, we care about your portfolio and your velocity

• Comfort using AI coding tools (Claude Code specifically) as a daily driver, not a novelty

• Architectural-level understanding of Kubernetes, AWS, GCP, and Azure; enough to read a cluster config, understand a VPC layout, parse IAM policies, and know what questions to ask. You don't need to be a DevOps engineer, but you need to speak the language

• Familiarity with IaC tooling: Terraform, Pulumi, or CloudFormation, as structured data sources you can reason over

Strong signals:

• Experience with Neo4j, AuraDB, or similar graph databases in production

• Background in static analysis, AST parsing, or code intelligence tooling

• Exposure to enterprise software environments with multiple repositories and complex dependency chains

• Comfort working asynchronously in a distributed team

How We Hire

No CV screening. No algorithm whiteboard.

1. Portfolio review: show us what you've built. GitHub repos, demos, write-ups. We want to see knowledge graph work and evidence of fast delivery

2. Live build session: a practical, time-boxed session where you build something real using Claude Code. We'll assess how you think, how you use the tools, and how you ship

3. Architecture conversation: a discussion about how you'd approach a real problem in our domain. No trick questions, just a conversation between engineers

Working Setup

• Remote-first, async-friendly

• Distributed team across Europe (Poland, Sweden, UK)

• High autonomy, high accountability

• Claude Code is the default development tool. If you have not used it yet, get comfortable before applying