Knowledge Engineer

Dyad is seeking a Knowledge Engineer to design, evolve, and operationalise the structured clinical knowledge that underpins our AI systems.

This is a mid-to-senior individual contributor role focused on classical knowledge engineering and graph data curation. The role combines deep semantic expertise with real-world accountability: ensuring that clinical and administrative concepts are modelled correctly, knowledge graphs remain reusable across customers, and ontologies scale without becoming brittle or bespoke.

You will play a critical role in making Dyad's clinical intelligence durable, reusable, and transparent - ensuring knowledge is embedded in structured systems rather than trapped in individual heads or ad-hoc code.

This role includes meaningful collaboration with customers and internal teams and is offered on a hybrid basis from our London office.

Core responsibilities Ontology & knowledge graph design

• Lead the design, extension, and maintenance of medical ontologies, terminologies, and knowledge graphs.
• Ensure semantic consistency and concept reuse across customers and products.
• Model clinical and administrative concepts in ways that support machine reasoning, not just human categorisation.
• Ensure interoperability with relevant standards (e.g. SNOMED CT) and alignment with real-world workflows such as US quality measures.

Customer-facing knowledge elicitation

• Engage directly with customers to understand source documents, coding practices, and workflow nuances.
• Translate customer feedback and AI failure modes into graph corrections, schema updates, or ontology extensions.
• Identify when system errors reflect knowledge modelling gaps rather than model performance issues.
• Act as a calm, credible interface between customer reality and system structure.
• Own the representation, versioning, and evolution of clinical concepts within Dyad's knowledge systems.
• Define provenance and lineage requirements for clinical data.
• Coordinate major ontology shifts with Product and AI leadership while retaining execution authority.
• Ensure knowledge artefacts are well-documented, transparent, and reusable across teams.

Automation & tooling

• Collaborate to design and build pipelines that ingest structured, semi-structured, or unstructured data into graph systems.
• Collaborate with engineers to automate ontology updates and reduce manual curation overhead.
• Improve tooling to prevent knowledge bottlenecks and re-centralisation.
• Ensure robust documentation and knowledge transfer to support adoption across the organisation.

Requirements

A minimum of a bachelor's degree in computer science with a focus on logical systems is required (or equivalent education), as well as at least 2 years commercial experience delivering knowledge-based systems in real production environments.

Semantic & ontology expertise

• Strong experience in ontology engineering and semantic technologies, including:
• OWL
• RDF
• SHACL
• Deep understanding of logical systems and knowledge modeling
• Experience with constraint validation and inference.
• Ability to design ontologies that are both expressive and operationally practical.
• Hands-on experience with triple stores such as:
• Stardog
• GraphDB
• Experience with graph databases (e.g. AWS Neptune) and hybrid storage models.
• Advanced SPARQL skills.
• Comfortable with SQL and NoSQL where appropriate.
• Experience building ingestion pipelines to populate graphs at scale.
• Ability to collaborate effectively with software engineers to operationalise semantic infrastructure.
• Strong appreciation for documentation, versioning, and knowledge durability.

Domain & operating context

• Experience working with healthcare data, clinical terminologies, or administrative coding systems is strongly preferred.
• Ability to engage confidently with customers and translate domain knowledge into structured representations.
• Comfortable working in a startup environment where ownership and initiative are expected.

Personal attributes

• Systems thinker with strong attention to conceptual clarity.
• Pragmatic rather than academically purist.
• Comfortable balancing theoretical correctness with real-world constraints.
• Motivated by building durable, reusable knowledge systems that improve AI reliability.

Our hiring process

• Introductory screening interview (30 minutes)
• Technical and domain interview with Applied AI and Product leadership
• Final interview and offer
• Company pension
• 25 days of paid annual leave (pro-rata)
• Flexible hybrid working environment
• Employee Assistance Programme
• Modern, dog-friendly office near Chancery Lane with free drinks

Dyad's mission is to improve the delivery and efficiency of healthcare.

We are building a platform to model and manage the flow of information within healthcare organisations, improving outcomes for patients, payers, and healthcare providers. We believe data handling in current healthcare systems is needlessly complex and disconnected, leading to isolated and inefficient decision making. To showcase how this technology can advance the delivery of healthcare and improve lives, we build and deploy products for healthcare providers and payers into the UK and US markets.

Dyad is an energetic, health-tech startup, currently around forty employees. Our team is growing as we explore new markets and opportunities. We are passionate about technology and its applications in worthwhile ventures. New joiners will have a significant impact on the direction of the company, as well as our culture.

Our products

Dyad's Platform: Dyad's products are founded upon our Semantic AI platform, which enables payers and providers to access cutting-edge AI capabilities for their own use cases and applications. Our partners either use the platform APIs directly or work with us to develop applications for their use cases. For more information, please see our Platform page.

Primary care operations : Dyad develops a suite of products for healthcare operations, including:

• BetterLetter, our AI tool helping practices decrease their admin burden in processing clinical letters. We use this to reduce staff time spent identifying codes to be applied to the record as well as suggesting follow-up tasks and workflow optimisations. BetterLetter helps providers save time, save cost, improve performance under audit and build staffing resilience.