Lead Specialist, AI Scientist

Pearson•Austin, TX

21h•$140,000 - $160,000•Remote

About The Position

Pearson is seeking a Lead Specialist, AI Data & Knowledge Architect to design and build the knowledge foundation for its next-generation intelligent K–12 assessment and learning systems. The position is part of the Measurement & Learning team and works across Content Development, Learning Science, Technology, AI Science, and Psychometrics, translating assessment and measurement expertise into extensible computational representations. Prior assessment experience is helpful but not required. This founding, hands-on role will develop a layered architecture of interconnected knowledge graphs beneath the assessment authoring system. The knowledge layer will represent standards, competencies, claims, evidence, task designs, scoring, and content in machine-readable forms to support AI agents and workflows, as well as formative and summative assessment. The architecture should support predictive, statistical, graph-based, deterministic, and generative AI methods. The role designs knowledge models and graph/data architecture; connects source systems; develops schemas, APIs, and services; and builds retrieval and context approaches that make structured knowledge useful to AI systems. It also establishes traceability, quality measures, and responsible data practices. The architect works closely with engineers, AI scientists, psychometricians, learning scientists, content teams, and subject-matter experts. This role requires autonomy, rapid prototyping, empirical testing, and clear communication of architectural decisions and tradeoffs.

Requirements

Advanced degree in Computer Science, Artificial Intelligence, Information Science, Knowledge Engineering, or a related field, or equivalent experience.
Demonstrated experience designing and building knowledge graphs in production or substantial applied settings.
Hands-on experience with knowledge modeling and semantic schema design using approaches such as ontologies, property-graph schemas, vocabularies, and taxonomies.
Strong command of graph and semantic technologies such as RDF, OWL, SHACL, SPARQL, and property graphs, including their architectural tradeoffs.
Strong data architecture skills, including data modeling, integration pipelines, identity, versioning, provenance, and lineage.
Strong programming ability in Python or comparable languages, with a record of building working systems and APIs.
Deep understanding of predictive and generative AI, including how retrieval-grounded and agentic systems use structured knowledge.
Track record of building novel systems where no established solution existed, with sound judgment and high autonomy.
Ability to translate expertise from other disciplines into computational representations and communicate tradeoffs clearly.

Nice To Haves

Strong understanding of educational knowledge structures, including K–12 standards, competency frameworks, skills taxonomies, or credential frameworks, ideally through modeling them as structured data.
Experience with knowledge graphs powering applications such as content generation, framework alignment, personalized guidance, or analytic insights.
Experience with formal ontology engineering, including OWL-based modeling, reasoning, or validation.
Experience building hybrid AI systems combining machine-learned or generative models with structured knowledge, rules, or constraints, including neuro-symbolic approaches.
Experience with graph machine learning, graph embeddings, link prediction, or graph neural networks.
Experience in domains where outputs must be validated against reliable evidence or expert judgment, such as scientific, healthcare, financial, or legal AI.
Experience with evaluation datasets, human annotation and adjudication workflows, or benchmark development.
Familiarity with educational measurement, psychometrics, principled assessment design, or assessment content systems.
Publications, patents, or open-source contributions in knowledge representation or graph technologies.

Responsibilities

Design and build the knowledge graph and knowledge model for assessment and learning domains, selecting appropriate ontologies, schemas, vocabularies, and taxonomies.
Represent standards, competencies, skills, claims, evidence, task characteristics, misconceptions, content, and relationships as computationally usable structures.
Design extensible models for content areas, programs, assessment contexts, and future AI capabilities.
Establish stable identity, versioning, provenance, lineage, and lifecycle status for knowledge objects; distinguish authoritative knowledge, expert judgments, derived information, and machine-generated hypotheses.
Evaluate and select graph and semantic technologies based on extensibility, interoperability, reasoning needs, maintainability, and downstream AI use.
Build pipelines, mappings, and synchronization processes connecting item banks, standards, metadata, psychometric data, and other sources.
Develop schemas, APIs, data contracts, and services for applications, agents, and analytic workflows to access the knowledge architecture.
Partner with software engineers to integrate the knowledge layer into the assessment authoring system.
Design graph, semantic, vector, and hybrid retrieval and context architectures for AI and machine-learning systems.
Create task-specific knowledge projections and maintain traceability between AI outputs and the knowledge used to produce them.
Work with Applied AI Systems Engineers to support reliable generative, predictive, agentic, and analytic capabilities over the shared knowledge layer.
Exercise autonomy in shaping architecture, prototype quickly, test assumptions empirically, and revise approaches when evidence warrants.
Evaluate emerging graph-based, knowledge-grounded, and neuro-symbolic approaches and adopt those that demonstrate measurable value.
Make architectural decisions, assumptions, and tradeoffs explicit and open to challenge.
Design knowledge systems for transparent, auditable, and traceable AI reasoning, with appropriate points for expert review and judgment.
Collaborate with AI engineers, psychometricians, and content experts to test whether knowledge and retrieval designs improve accuracy and reliability.
Establish quality, coverage, and consistency measures for knowledge assets and uphold responsible practices for privacy, security, access control, and appropriate use of student and content data.
Partner with Technology teams, software engineers, AI Science colleagues, Applied AI Systems Engineers, psychometricians, learning scientists, content development teams, and subject-matter experts to co-design and implement the architecture.
Communicate complex architectural concepts clearly to technical and non-technical audiences.