Applied AI engineer, knowledge graph

Adverant, Inc.; Dublin; Hybrid; EUR 70,000 - 110,000; Full time

Job summary

Adverant Limited is building the heart of Nexus—our knowledge graph, ontology builder and GraphRAG retrieval—so every AI answer has full contextual sources across a company's knowledge base.

You will work across the stack from contract-to-facts graph creation to delivering model inputs with source attachments, shaping how information is retrieved and interpreted for trustworthy results.

Qualifications

• Production experience with retrieval, search or knowledge graphs.
• Strong Python, and comfort with a typed language.
• Experience evaluating language models on real tasks.
• A habit of checking results against what people actually need.
• Clear writing about trade-offs.

Responsibilities

• Improve extraction, entity resolution and the way facts keep their sources.
• Build and tune GraphRAG retrieval, including community summaries for broad questions.
• Make the ontology builder easier for domain experts to use.
• Measure quality with real test sets, and publish what you find internally.
• Work with industry leads to model new domains.

Skills

Retrieval systems

Knowledge graphs

Python

Typed languages

LM evaluation

Trade-off writing

Tools

Graph databases

Ontologies

Document parsing

Job description

Build the heart of Nexus: the knowledge graph, the ontology builder and the GraphRAG retrieval that gives every AI answer the context of a whole business.

Our customers want answers they can trust, drawn from everything their company knows. That depends on extraction that gets facts right, an ontology people can shape on screen, and retrieval that follows the connections between things rather than matching words.

You will work across the stack, from how a scanned contract becomes facts in the graph to how those facts reach a model with their sources attached.

What you'll do

• Improve extraction, entity resolution and the way facts keep their sources.
• Build and tune GraphRAG retrieval, including community summaries for broad questions.
• Make the ontology builder easier for domain experts to use.
• Measure quality with real test sets, and publish what you find internally.
• Work with industry leads to model new domains.

What you bring

• Production experience with retrieval, search or knowledge graphs.
• Strong Python, and comfort with a typed language.
• Experience evaluating language models on real tasks.
• A habit of checking results against what people actually need.
• Clear writing about trade-offs.

Good to have

• Graph databases, ontologies or semantic modelling.
• Information extraction from messy documents and scans.
• Work in insurance, law, finance or another document-heavy field.

What we offer

Real ownership of a product used in serious places, a team that writes things down, and the time to do the work properly. The written offer sets out pay, level and benefits.

How we hire for this role

• A first call with the hiring manager, about thirty minutes.
• Your work. Walk us through something you built, or take a short exercise.
• Meet the team in a working session on a real problem.
• An offer, in writing.

We give you an answer at every step.