Knowledge Architect (Taxonomy & Ontology)

Salesforce•San Francisco, CA

About The Position

Salesforce is seeking an experienced Taxonomist and Knowledge Architect to build and govern the taxonomy and ontology foundation for AgentExchange, an intelligent discovery and distribution platform for applications, agents, skills, MCP servers, APIs, and other capabilities. The role’s first priority is to deliver the capability ontology and semantic foundation for the capability reasoning engine, working closely with Search, Data Science/ML, Product, and other teams. The role will then extend that foundation to additional platform taxonomies. The Knowledge Architect owns semantic architecture supporting AI-powered discovery, capability matching, recommendations, and agentic workflows. This includes defining knowledge models, metadata schemas, controlled vocabularies, relationships, and machine-readable contracts that bridge customer intent and machine-actionable metadata. The role also develops governance, quality, verification, and evolution processes for the platform’s knowledge models and owns the taxonomy development roadmap. The successful candidate will bring expertise in information and knowledge architecture, semantic technologies, taxonomies, and AI systems, along with product judgment and experience delivering scalable platform capabilities. The position is a cross-functional product and architecture role, collaborating on technical design without serving as the primary implementation engineer.

Requirements

  • At least 5 years of experience in knowledge architecture, information architecture, semantic technologies, or related enterprise platform disciplines.
  • Experience building or leading products involving taxonomies, ontologies, knowledge graphs, semantic search, recommendation systems, or AI-powered discovery.
  • Strong understanding of taxonomy design, metadata modeling, entity relationships, knowledge representation, and information retrieval.
  • Ability to design faceted classification and model typed relationships beyond broader/narrower links.
  • Experience translating complex technical architectures into product strategies, requirements, and measurable outcomes.
  • Working knowledge of modern AI architectures, including LLMs, RAG, embeddings, and structured reasoning.
  • Experience working closely with search, data science, or ML teams and reviewing model or search outputs to iterate with those teams.
  • Comfort with SQL, scripting, or notebooks, and with writing requirements against schemas and APIs.
  • Experience driving cross-functional initiatives in complex enterprise organizations and collaborating with engineers and architects on technical design decisions without serving as the primary implementation engineer.

Nice To Haves

  • Master’s degree in Library and Information Science (MLIS), Information Architecture, Computational Linguistics, or a related discipline.
  • Experience with RDF, OWL, SKOS, SHACL, schema.org, or related semantic modeling standards.
  • Experience with AI agents, MCP, tool discovery, capability registries, or agent orchestration.
  • Experience designing taxonomies or knowledge models for large-scale marketplaces, developer platforms, or enterprise software ecosystems.
  • Experience building taxonomies or knowledge models that feed retrieval or RAG systems.

Responsibilities

  • Define the canonical knowledge model, taxonomies, metadata schemas, ontologies, and relationships for applications, agents, skills, MCP servers, APIs, and their capabilities.
  • Model dependencies, prerequisites, compatibility, constraints, and composability across capabilities, including how the platform evaluates compatibility and combinations.
  • Define attributes and relationships that enable agents to evaluate whether a capability is trustworthy, compatible, authorized, and appropriate for a task, partnering with the reasoning engine team, Security, and Trust.
  • Map customer intent, business problems, use cases, and natural-language queries to available capabilities.
  • Define controlled vocabularies and facets, including industry, cloud, edition, persona, compliance requirements, and deployment type, and describe native and partner capabilities in comparable terms.
  • Co-own ground truth and evaluation with Data Science/ML and Search, including labeled examples, edge cases, and quality definitions for matching accuracy, recommendation quality, and reasoning correctness.
  • Design explainability and confidence into the knowledge model so customers and agents can understand recommendations.
  • Enable AI assistants and agents to discover and consume AgentExchange capabilities through APIs, MCP interfaces, and other machine-readable contracts.
  • Partner with Engineering, Data, AI, Design, Marketplace, and other teams on knowledge graphs, semantic retrieval, reasoning approaches, and agentic discovery; deliver reusable platform capabilities.
  • Write build-ready requirements and versioned schemas for engineers.
  • Develop extensible semantic structures for new asset types and the evolving agentic ecosystem.
  • Own AgentExchange information architecture strategy across classification, metadata, content relationships, and findability, maintaining consistent standards across first-party and third-party catalogs.
  • Partner with Search, Personalization, and UX teams to improve navigation, filtering, semantic search, and recommendations.
  • Redesign listing metadata over time to provide structured, consistent, machine-readable descriptions, including functionality, inputs and outputs, prerequisites, permissions, dependencies, compatibility, limitations, and supported use cases.
  • Establish knowledge-model governance, including a shared glossary, named owners for terms and data sources, decision rights, and field-level source-of-truth indicators.
  • Set metadata quality, freshness, completeness, and verification standards, distinguishing publisher-provided claims from independently validated attributes.
  • Define how trust and verification signals are represented and exposed through structured APIs and MCP interfaces, in partnership with teams that own the trust framework.
  • Define provenance and refresh rules for outcome and adoption signals such as success rates, effort estimates, and co-install affinity.
  • Develop scalable ontology-evolution and catalog-consistency processes that maintain accuracy, extensibility, and interoperability without creating unnecessary publishing friction.
  • Own the taxonomy development roadmap, prioritizing the capability reasoning engine and conducting discovery with customers, partners, developers, and internal Salesforce product teams.

Benefits

  • Time off programs
  • Medical insurance
  • Dental insurance
  • Vision insurance
  • Mental health support
  • Paid parental leave
  • Life insurance
  • Disability insurance
  • 401(k)
  • Employee stock purchasing program
  • Certain roles may be eligible for incentive compensation and equity.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service