---
id: "action-knowledge-retrieval"
type: "action-item"
source_timestamps: ["§ Myth 4"]
tags: ["customer-service", "knowledge-base"]
related: ["concept-unstructured-data-leverage", "contrarian-messy-data"]
action: "Connect LLMs to existing unstructured product manuals to build a rapid knowledge retrieval system."
outcome: "Enable agents to diagnose and resolve customer issues up to 10 times faster."
speakers: ["Doug J. Chung", "Candace Lun Plotkin", "Siamak Sarvari", "Jennifer Stanley", "Maria Valdivieso"]
sources: ["attention"]
sourceVaultSlug: "hbr-seg-attention"
originDay: 4
articleStem: "hbr-cl-90-genai-myths-sales-marketing"
sourceUrl: "https://hbr.org/2025/02/5-gen-ai-myths-holding-sales-and-marketing-teams-back"
sourceTitle: "5 Gen AI Myths Holding Sales and Marketing Teams Back"
---
# Build Unstructured Knowledge Retrieval

## Action: Build Unstructured Knowledge Retrieval

**Do this:** Instead of waiting to clean databases, point publicly available **LLMs** at existing unstructured internal materials — product manuals, PDFs, troubleshooting Q&A documents — to assist customer-service agents.

**Expected outcome:** Enable agents to diagnose and resolve customer issues **up to 10 times faster** (the global machinery distributor case).

**Myth addressed:** Myth 4. See [[concept-unstructured-data-leverage]] and the contrarian framing [[contrarian-messy-data]]. In practice this is typically built with **retrieval-augmented generation (RAG)** — keep chunking, metadata, and access control in scope.