Adventures in setting up a knowledge system for a research group
summary
A research group builds a shared knowledge system using Confluence + Bitbucket + Jira, migrates all code and standardizes project spaces, and defines measurable outputs like one knowledge product per 300 hours (proof-of-concept demos, white papers, code PRs, logbooks). The knowledge-management team dissolves after rollout but the structured Confluence spaces stick, adoption grows unevenly across ~20 concurrent projects, and the setup pays off for handovers even though full product-generation targets lag.
AI-written summary · about
chapters (25)
- 0:00Introduction
- 1:03Who am I
- 1:56About the group
- 2:56How we work
- 4:11Collaborative robots
- 4:26Drones
- 4:59Projects
- 5:44Knowledge and information
- 6:29Knowledge and education
- 7:00Storage solutions
- 7:57The problem
- 10:15Lean
- 10:31Fishbone diagram
- 11:31Why not use a standard workflow
- 12:29What are knowledge products
- 13:05Scoring knowledge products
- 14:34Implementation modes
- 16:41Organization
- 18:08Procedures
- 18:38The abyss
- 19:24Where to store it
- 25:11Write to Docs
- 25:52Disagreements
- 29:49Tools
- 32:21Knowledge Products
related talks
24mHow did we open source our knowledge and practices
40mAll together now: Turning technologists into documentarians
21mArchitecting a Better Knowledge Base
17mCreating a knowledge-sharing culture
23mKnowledge sharing in the company
35mKanban, our way
39mOrganizing a Confluence hoard, or, does this page spark joy?
34mScaling Silos: Distributed Knowledge