Skills related to the design, management, and continuous improvement of enterprise retrieval systems — including both traditional search and AI-driven semantic retrieval. This covers index tuning, metadata optimisation, embedding-based search, chunking strategies, RAG corpus curation, query interpretation, and the governance of authoritative sources. The discipline ensures that users and AI systems can reliably find, interpret, and synthesise organisational knowledge.
Understands how content structure, metadata, and authority affect search and AI retrieval. Applies templates, tagging standards, and basic semantic enrichment to improve findability. Uses AI-assisted tools to generate summaries, metadata, and classifications. Identifies simple retrieval issues (poor titles, missing metadata, duplicated sources) and escalates complex ones.
Configures and tunes retrieval systems using relevance weighting, semantic enrichment, metadata governance, and AI-powered indexing (e.g., auto-tagging, embeddings). Implements RAG enhancements such as chunking optimisation, canonical source prioritisation, and similarity-search tuning. Analyses search logs and AI query behaviour to diagnose failures and recommend corrections to content or metadata.
Designs and governs the enterprise retrieval architecture across search, semantic indexing, and AI ingestion pipelines. Manages authority, provenance, and corpus composition to reduce equivocality and hallucination. Oversees RAG tuning, embedding strategies, vector store optimisation, and retrieval evaluation frameworks (precision, recall, attribution). Influences product teams and content owners to create AI-legible, structurally sound, authoritative content. Leads telemetry-driven improvement across search and AI experiences.
We have hopefully created this competency matrix with thought and care. It is not the only competency framework in the world, and relates specifically to the intranet and digital workplance profession. It therefore concentrates on some things and ignores others.
If you find an error, disagree wholeheartly or feel there is a glaring ommission we'd love to know.
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.