Enterprise context and search is the layer that decides what an agent is allowed to know, what it can actually retrieve, and whether the answer it gives can be traced back to something the company still stands behind.
Library: 20 published articles · Target URL: /blog/enterprise-context-and-search · Last verified: 2026-10-19
What this cluster covers
This cluster owns the knowledge substrate underneath agent work: how company content is connected and indexed, how permissions survive retrieval, how relevance is ranked and cited, how entity and context models give an agent something coherent to reason over, and how a team evaluates the platforms that promise all of it — Glean and its alternatives among them.
Where the boundary sits.
MCP Fundamentals owns how an agent connects to data and tools. This cluster owns what it gets, and what it is allowed to get, once connected.
Human–Agent Collaboration owns who works with whom and under what authority. This cluster owns the knowledge they work from.
Comparisons & Alternatives prices the products. This cluster tests whether the retrieval and governance model underneath them holds up.
Agentic Research, GTM & AI Visibility applies these systems inside one function. This cluster builds and evaluates the system itself.
The 20 published articles
Every link below is live on the public blog; the sentence after it says which job the article does inside the cluster.
What enterprise search is (3)
What Is Enterprise Search? Architecture, Permissions, and Use Cases — the foundation article: connected sources, permission-preserving retrieval, ranked evidence, cited answers, and how a team should evaluate it.
AI Enterprise Search: How Grounded Answers Work Across Company Data — the AI layer on top of the foundation: identity controls, retrieval lanes, ranked evidence, and verifiable citations.
Workplace Search vs Enterprise Search: What Teams Actually Need — the distinction teams get wrong when they compare vendors: employee-experience search against configurable, owned search infrastructure.
Architecture and security (3)
Enterprise Search Architecture: Connectors, Indexes, ACLs, and Answers — the production shape: connectors, parsing, hybrid indexes, ACL security trimming, grounded answers, freshness, observability, and evaluation.
Secure Enterprise Search: Permission-Aware Retrieval and Audit — the control article: trusted identity, source ACLs, query-time trimming, safe caches, revocation proof, prompt-injection controls, and audit evidence.
Hybrid Search Explained: Keyword, Vector, and Graph Retrieval — why one retrieval signal is not enough, and how rank fusion, ACL filtering, reranking, and evaluation fit together.
Retrieval and reasoning patterns (4)
What Is Agentic RAG? Architecture, Benefits, and Failure Modes — retrieval an agent plans and evaluates itself, with loops, latency, cost, and production failure modes made explicit.
RAG Architecture for Enterprise Knowledge Systems — the enterprise design guide: ingestion, hybrid retrieval, permissions, citations, evaluation, freshness, deletion, and repair.
What Is a Context Graph? A Practical Model for AI Workspaces — who and what exists, how it relates, what happened, and under which permissions — and how it differs from a knowledge graph.
Context Engineering vs Prompt Engineering: What Changes in Production — the shift from phrasing an instruction to assembling evidence, tools, state, memory, and permissions.
Knowledge models and governance (3)
Enterprise Knowledge Base: Architecture, Governance, and Ownership — knowledge treated as an operating system: source authority, ownership, permissions, lifecycle governance, and operations.
Enterprise Knowledge Graph: Entities, Permissions, and Provenance — modelling entities and relationships without losing permissions, provenance, time, or source authority.
AI Knowledge Management: From Retrieval to Reusable Work — the loop that turns retrieval into work both a person and an agent can reuse.
Evaluating and buying (6)
Best Enterprise Search Software: An Evaluation Framework — a buyer framework that starts from permissions, connectors, answer quality, and pilot evidence rather than connector counts.
Best Glean Alternatives for Enterprise Knowledge and AI Agents — what you are actually replacing when you replace Glean, and which route fits which job.
Glean Pricing Explained: Licenses, AI Usage, and TCO — why there is no public per-seat list, and how to model FlexCredits, connectors, implementation, and three-year cost.
Best Knowledge Management Software for AI-Enabled Teams — the storage question replaced by the governance, search, lifecycle, and agent-readiness question.
Enterprise AI Assistant: Capabilities, Controls, and Evaluation — how to judge an assistant by knowledge, citations, actions, permissions, operations, and cost per successful task.
AI Agent Builder: What to Evaluate Before You Buy — the build-versus-buy checklist: knowledge, tools, identity, testing, lifecycle, observability, human control, and production risk.
Agent execution (1)
AI Agent Orchestration: Architecture, State, and Human Control — how a goal becomes a bounded sequence of agent, model, retrieval, and tool steps with checkpoints, retries, and human approval.
Evaluation checklist
Before adopting anything in this cluster, ask:
Can every retrieved passage, snippet, and generated answer be traced to an authorised source?
Do permissions survive retrieval, caching, and revocation in production — not only in a demo?
When retrieval is wrong, is the failure visible, bounded, and repairable?
Which content is stale, superseded, or owned by nobody, and who resolves that?
Can a person see why an answer was produced before acting on it?
What is the cost per successful task, not per seat?
Related pillars
Sources and freshness
Every article in this cluster carries its own date on its own page. This hub was checked against the live public blog on 2026-10-19: all 20 URLs returned HTTP 200 and all 20 are listed in the blog index at /pub/blog.