Knowledge Architect

Must Have Skils: Knowledge Graphs & AI Semantics, Investment & Financial Services domain

Job Description

We are hiring a Lead Knowledge Architect for an investment and financial-services client whose departments (front office, risk, finance, operations, compliance) describe the same concepts in different vocabularies, causing their AI systems to return inconsistent and uncertain answers. You will build the unified semantic foundation (ontologies, taxonomies, knowledge graphs) that reconciles these vocabularies and tune how AI retrieves and reasons over it so answers become consistent, explainable, and trustworthy. This requires a seasoned data architect with deep graph expertise and enough investment-business fluency to settle semantic disputes between a portfolio manager, a risk officer, and a controller.

What you will do:

• Design and own the enterprise ontology and knowledge graph for instruments, issuers, counterparties, portfolios, positions, transactions, and reference data, aligned with standards such as FIBO, ISO 20022, and LEI/ISIN.
• Lead cross-departmental vocabulary reconciliation: define canonical concepts with lineage to each department's usage and run a governed business glossary and semantic layer.
• Architect, optimize, and benchmark graph databases and query layers (Neo4j, Amazon Neptune, TigerGraph, RDF stores; SPARQL/Cypher/Gremlin) at financial-data scale.
• Optimize AI outputs through Graph RAG and hybrid retrieval, entity resolution, grounding and citation strategies, and evaluation frameworks measuring consistency, accuracy, and uncertainty.
• Build knowledge-extraction pipelines from structured (OMS/PMS, accounting, risk, market data) and unstructured sources (research, filings, contracts) with lineage and quality controls, meeting regulatory and auditability requirements.
• Act as the client's semantic authority: facilitate SME workshops, present to senior stakeholders, and mentor engineers and analysts.

What we are looking for:

• 8+ years in data architecture, including 4+ years designing and operating production knowledge graphs or semantic platforms.
• Deep hands-on graph expertise: property-graph and RDF/OWL modeling, SHACL, SPARQL, Cypher, Gremlin, performance tuning, and at least two major graph databases at scale.
• Strong investment and financial domain knowledge: asset classes and instrument lifecycles, portfolio and position data, risk and performance measurement, fund accounting, trade lifecycle, and regulatory context (MiFID II, BCBS 239, SEC/ESMA reporting). Prior work at an asset manager, bank, hedge fund, or financial data vendor strongly preferred.
• Proven experience improving LLM output quality via semantic grounding: RAG/GraphRAG, embeddings and vector search, entity resolution, retrieval evaluation, and hallucination measurement.
• Experience with data catalogs and glossary tooling (Collibra, Alation, Atlan) and master/reference data management; proficiency in Python and SQL and modern data platforms (Spark, Databricks, Snowflake).
• Exceptional facilitation and communication skills; able to align competing stakeholders and translate outcomes into formal models.