> Markdown version of [/jobs/ext/3322940-director-knowledge-graph-and-entity-infrastructure](https://www.wearedevelopers.com/jobs/ext/3322940-director-knowledge-graph-and-entity-infrastructure). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # Director, Knowledge Graph and Entity Infrastructure - **Company:** Fitch Group - **Location:** London, UK - **Contract:** Permanent contract - **Skills:** Artificial Intelligence, Data Analysis, Information Engineering, Data Sharing, Query Languages, Graph Database, Python (Programming Language), Linked Data, SPARQL, SQL Databases - **Published:** September 24, 2026 - **Apply:** https://www.indeed.com/viewjob?jk=b2ba28a536b2a92d ## About the Role Strong candidates may come from knowledge graph engineering, data engineering, applied data science or semantic technology. We care more about what you have built and how you approach a problem than a particular tool stack or formal ontology background. * You have built a knowledge graph or another connected data application for analytics, content or AI, and can take an idea from domain conversations to a working implementation. * You are a hands-on builder with Python and SQL who can own data mappings, graph queries, tests and the first usable version of a solution. * You can model real-world relationships with appropriate attention to identity, time, provenance and validation, adding detail as the use case requires it. * You have experience with a graph database or query language, or closely related semantic tooling. Familiarity with Cypher, SPARQL, RDF or property graphs is valuable; no single stack is required. * You can work with analysts, product managers, data scientists and engineers to define an ambiguous problem, make practical tradeoffs and deliver something other teams can use. What Would Make You Stand Out: * Knowledge of credit, financial markets, industry or country risk data, or experience making complex research content usable as structured information. * Experience applying graph-derived features to modeling, or connected retrieval to source-backed AI applications. * Experience with content enrichment, taxonomy design, temporal data, provenance or linked data standards across multiple teams or source systems. ## Description This role has the same dual remit: help selected initiatives deliver working knowledge graph applications, and establish the models, components and practices that future initiatives can adopt. You will work with data scientists, product teams, analysts, AI teams and the Core groups responsible for shared data and platforms. You will also help align our work with broader Fitch initiatives that connect research and intelligence across businesses. How You'll Make an Impact: You will lead the design and initial implementation of connected data and knowledge graph capabilities that make analytics easier to discover, combine and apply. Starting with a real product problem, you will work with subject experts to model the relevant concepts and relationships, write the code and queries, and put a working solution in users' hands. You will partner with engineering teams to make useful solutions durable and scalable. The model should be clear enough to extend, and it should evolve as real use cases reveal what is needed. Research Content and AI * Partner with research and content teams on tagging and enrichment that connects documents, topics, entities, events and analytical capabilities. Help make authored research easier to discover and reuse while preserving its source and context. * Contribute the semantic and graph modeling expertise needed for new product initiatives, which aim to connect intelligence across Fitch Ratings, BMI, Sustainable Fitch and CreditSights. Align definitions and interfaces with the wider program as its scope develops. * Work with AI teams to evaluate where connected context improves retrieval, grounded answers or other product workflows. Design for traceable sources and assess answer quality against real questions. Connected Analytics and Risk * Translate the unified analytics framework into a practical semantic model. Define analytical domains, subjects, lenses, capabilities, methods, inputs, outputs and the relationships among them so teams can understand and reuse work across products. * Model the connections that can explain risk pathways, such as how a country or industry development may relate to companies, exposures and credit outcomes. Make the meaning, timing, source and limits of each relationship explicit. * Work with data scientists to test whether graph relationships, features or algorithms improve a priority model or analytic use case. Compare the result with a simpler baseline and retain what proves useful. Implementation and Partnership * Embed in selected initiatives and personally build initial graph models, data mappings, queries and working implementations. With engineering and Core platform teams, develop the ingestion, validation and interfaces needed to make successful solutions reliable. * Use shared entity and identifier services as they mature. When a use case exposes a genuine linking problem, work with the teams that own entity resolution to define requirements and test an appropriately simple solution. * Establish reusable modeling patterns, examples and quality measures for relationships, provenance, temporal context and coverage. Teach partner teams to use and extend the capability, and help shape a larger specialist team if demand supports it. You First Year: Initial priorities will be chosen in collaboration with the Head of Analytics and product teams based on client value and available data. Early outcomes should include: * A working implementation for one priority analytics or risk question, using real data and a model validated with domain experts and prospective users. * A test of at least one adjacent use case, such as graph-derived features for a model or connected research for source-backed AI, measured against a simpler approach. * Reusable patterns, working examples and a practical adoption plan so future squads can apply what proves useful, with clear responsibilities across the Data Science Engine, Core platform and content teams.