Knowledge at Work

Thinking on how teams store, share, and synthesize what they know.

Abstract overlapping speech bubbles representing repetitive team communication
Featured · · 7 min read · Aisha Rahman

The Recurring Slack Question: Why Teams Keep Deriving the Same Answers

Every engineering team has a version of the same pattern: a question hits Slack, three senior engineers each share a different document, nobody consolidates, and next month someone asks again.
Abstract evolution from scattered documents to organized luminous structure
· 9 min · Priya Okonkwo

Knowledge Management in 2025: Why Static Wikis Are No Longer Enough

The wiki was invented to capture knowledge. But the way teams create and consume knowledge has changed.

 Abstract layered architecture representing document retrieval pipeline
· 10 min · Marcus Delacroix

What RAG Actually Means for Your Internal Documentation

Retrieval-Augmented Generation is a phrase you've probably heard. Here's what it actually looks like when applied to a real company's internal knowledge.

 Complex organic network graph with nodes and edges in amber and forest green
· 11 min · Marcus Delacroix

Building a Knowledge Graph from Scratch, Using the Docs You Already Have

Knowledge graphs sound like infrastructure projects. But for most teams the raw materials already exist in Confluence, Jira, and Slack.

 Magnifying glass with amber glow over blurred document background
· 8 min · Priya Okonkwo

What Confluence Search Misses (and Why It's Not Confluence's Fault)

Confluence search is designed to find documents. It's not designed to answer questions. That distinction explains most of the frustration.

 Abstract walled-off geometric shapes in dark tones representing isolated knowledge silos
· 8 min · Aisha Rahman

The Real Cost of Knowledge Silos, and Why It Compounds as You Hire

Every hire adds knowledge to your organization. Without synthesis infrastructure, the percentage accessible decreases with headcount.

 Thin glowing lines connecting floating geometric shapes representing source attribution
· 7 min · Marcus Delacroix

Why Source Attribution Matters More than Answer Quality in AI Knowledge Tools

A confident-sounding answer without a source is actually worse than no answer.

 Stacked layers of papers with one amber-colored card lifted from the middle
· 6 min · Theo Bergmann

The Answer Was in the Ticket: Why Jira Is a Hidden Knowledge Base

The decision to use gRPC over REST is somewhere in a ticket comment from 2024. These aren't edge cases.

 Abstract network of interconnected nodes representing semantic connections
· 9 min · Marcus Delacroix

Semantic Search Isn't Just Better Keyword Search, It's a Different Model of Retrieval

Teams come to Corpussage expecting faster keyword lookup. They stay because semantic retrieval surfaces answers they didn't know to search for.

 Empty modern desk with laptop and small plant in morning light
· 7 min · Priya Okonkwo

From First Day to First Merge: How Knowledge Access Changes Onboarding Time

The biggest variable in engineering onboarding time is how quickly a new hire can answer questions independently.

 Conceptual art of two abstract paths representing search diverging and synthesis converging
· 8 min · Aisha Rahman

Search vs. Synthesis: What Your Knowledge Tool Should Actually Do

The knowledge management space is broadly organized around search. Corpussage is organized around synthesis.

 Stacked document pages gradually fading from front to back
· 8 min · Marcus Delacroix

Knowledge Half-Life: Why Documentation Goes Stale Faster Than You Think

Documentation that was accurate six months ago is already partially obsolete. Most teams have no mechanism for knowing which parts are stale.