Ontology Engineer-Knowledge Graph & Identity

What You'll Do:

Ontology Implementation & Validation

• Implement and extend Samba's RDF/RDFS/OWL ontology schemas in the graph database - adding entity classes, properties, and constraints in a consistent, governed way under the direction of the Senior Ontologist

• Build and maintain SHACL validation shapes for post-load graph consistency checks; identify and triage data quality and schema violations

• Support ontology versioning, change log documentation, and consistency checking across schema updates

• Write efficient, well-structured SPARQL queries and graph traversals to support downstream data science and product use cases

Event-to-Ontology Derivation Pipelines

• Contribute to the event-to-ontology transformation and derivation layer - building PySpark/Databricks pipelines that aggregate raw TV viewership and web activity events into durable graph attributes (genre affinity, brand affinity, topic affinity, viewing summaries, lifecycle signals)

• Implement derivation logic specified by the Senior Ontologist and data science team; validate outputs against SHACL shapes before graph load

• Support incremental refresh and update logic aligned with the graph's batch refresh cadence

Technical Contribution

• Write production-quality Python - clean, well-tested, documented, and reusable by teammates

• Work with PySpark and Databricks to process and transform high-volume data as part of graph pipeline development

• Apply embedding-based approaches (semantic similarity, vector search) to entity matching and ontology alignment tasks

• Contribute to team tooling, documentation, and reusable components that improve knowledge graph development efficiency

Collaboration & Growth

• Partner closely with data engineering on pipeline design, data quality, and incremental ingestion patterns feeding the materialized graph substrate

• Participate in ontology design reviews and cross-functional working groups

• Work with product and operations teams to understand use case requirements and translate them into graph schema updates

• Actively develop expertise in W3C semantic web standards, RDF-native graph databases, and entity resolution under the guidance of the Senior Ontologist

Who You Are:

Must-Haves

• 2–4 years of hands-on experience in knowledge graph development, semantic data modeling, ontology engineering, or a closely related field

• Working knowledge of W3C semantic web standards: RDF, RDFS, OWL, and SPARQL - with practical experience querying or building in at least one triplestore or graph database

• Familiarity with SHACL or equivalent constraint and validation frameworks for graph data quality

• Strong Python skills - clean, readable, production-quality code with testing and documentation

• Solid understanding of data modeling fundamentals - entity-relationship design, taxonomies, hierarchies, and how to represent complex real-world relationships in structured form

• Familiarity with entity resolution or data matching concepts - understanding of why the same real-world entity appears under different identifiers across data sources

• Bachelor's degree required in Computer Science, Information Science, Mathematics, or a related field; Master's preferred

• Detail-oriented and proactive about flagging data quality issues and schema inconsistencies

Strongly Preferred

• Hands-on experience with Amazon Neptune or Stardog - or equivalent RDF-native triplestore; exposure to data virtualization (Neptune Orion or Stardog Virtual Graphs) a plus

• Working knowledge of PySpark and Databricks - particularly for large-scale event aggregation and transformation pipelines

• Familiarity with embedding models, vector search, or semantic similarity - applied to entity matching, ontology alignment, or knowledge graph enrichment

• Experience with LLM APIs or RAG-based approaches applied to information extraction, entity disambiguation, or schema mapping

• Domain knowledge in media, entertainment, or ad tech - content metadata, advertising entities, TV viewership data, or audience/identity data

• Exposure to identity resolution, probabilistic record linkage, or device graph approaches