Intro an Enterprise Search Tool

practices/implement-a-documentation-search-engine.md

@@ -0,0 +1,55 @@
+# Implement a Document Search Engine
+
+When critical system knowledge is locked away in the minds of a few tenured employees, bottlenecks are created, onboarding slows, and reliance on key personnel increases. Incomplete or outdated context will always lead to rework. An enterprise search tool—such as [Glean](https://www.glean.com/) or [Claude](https://claude.ai/) for Enterprise—unifies access to information across JIRA, Confluence, SVN, shared drives, and communication platforms.
+
+This practice allows teams to quickly find what they need without interrupting others or digging through multiple systems that require specialized knowledge. By centralizing institutional knowledge and making it accessible, enterprise search boosts team autonomy, reduces wasted time, and supports a healthier learning culture where answers are self-served rather than gatekept.
+
+## When to Experiment
+
+- You are a new developer and you need to learn how to find accurate, updated documentation so you can onboard quickly without constantly interrupting mentors.
+- You are a an engineer and you need to ensure you have quick access to system knowledge to avoid wasting time sifting through multiple systems during testing.
+- You are a product owner or developer and you need access to complete and up-to-date requirements to produce quality tickets and avoid downstream rejections.
+- You are a senior engineer and you need to ensure that specialized knowledge is accessible to the wider team so you can reduce interruptions and boost team autonomy.
+
+## How to Gain Traction
+
+Implementing a document search engine requires more than just installing software; it requires mapping your knowledge ecosystem and training the team on how to retrieve it.
+
+### Map the Knowledge Silos
+
+Identify the highest-traffic repositories of knowledge that are currently disconnected. Usually, this begins with your ticket tracking system (e.g., [JIRA](https://www.atlassian.com/software/jira)), your documentation hub (e.g., [Confluence](https://www.atlassian.com/software/confluence)), and your version control system. Audit these sources to ensure permissions are clean before connecting a search tool, as effective search will surface documents that were previously "security through obscurity."
+
+### Connect and Pilot
+
+Select an enterprise search tool and connect it to your two most critical data sources. Roll this out to a small pilot group—specifically targeting new hires and senior leads who feel the burden of questions most. Use this phase to tune the search relevance and ensure that the tool is indexing metadata correctly so that results are ranked by recency and relevance.
+
+### Establish "Search First" Protocols
+
+To drive adoption, the team must shift from an "Ask First" to a "Search First" culture. Encourage senior staff to respond to questions with links to the search result rather than typing out the answer again. If a search comes up empty, use that as a trigger event to create the missing documentation immediately, ensuring the next search yields a result.
+
+## Lessons From The Field
+
+- _Reliance on Language Models for Security_ - Teams often realize their permission settings are lax only after a search engine surfaces sensitive HR or roadmap documents to the whole engineering org. Make sure security permissions are handled with the appropriate data layer security and not through system prompts or other language model means.
+- _Trust decay from stale data_ - If the top three search results are deprecated documents from three years ago, users will quickly abandon the tool. You must archive old data or boost the ranking of fresh content.
+
+## Deciding to Polish or Pitch
+
+After experimenting with this practice for **4-6 weeks,** bring the team together to determine whether the following metrics and/or signals have changed in a positive direction:
+
+### Fast & Intangible
+
+**Reduced Context-gathering Delays.** Feedback from developers should indicate less frustration and time spent "hunting" for requirements or historical context before starting a ticket.
+
+## Supported Capabilities
+
+### [AI-accessible Internal Data](/capabilities/ai-accessable-internal-data.md)
+
+This practice is one option for implementing AI-accessible Internal Data. The other option is building the system your self with in-house developers. However, you will likely only see the benefits of this if you are a large enough company with complex enough needs to justify putting a team of developers toward it. The vast majority of companies should opt for an off the shelf solution.
+
+### [Learning Culture](/capabilities/learning-culture.md)
+
+By centralizing institutional knowledge and making it accessible, enterprise search boosts team autonomy, reduces wasted time, and supports a healthier learning culture.
+
+### [Documentation Quality](/capabilities/documentation-quality.md)
+
+Excellent documentation is accurate, clear, complete, and accessible. This practice ensures that high-quality documentation is actually found and used, enabling teams to make informed decisions.

templates/new-practice.md

@@ -1,4 +1,4 @@
-# Use Data-generation Tools
+# Practice Name
 
 Introduction. 2-4 paragraphs.