Organizational knowledge tracker for the AI-generated code era
Problem
With AI producing code faster than teams can absorb it, organizations are losing track of context, what documentation is out of date, and what the source of truth is. Nobody holds deep domain expertise on any single system anymore, and finding who to ask about a given area is becoming impossible.
Opportunity
A tool that continuously maps an organization's knowledge: which docs are stale, who touched what last, who holds implicit expertise on each system — a living expertise directory synced to the codebase.
Market analysis
Strong timing: AI-generated code is making the 'who do I ask' problem measurably worse, and enterprises already pay heavily for adjacent solutions. But Glean's knowledge graph explicitly maps who the experts are, so a solo builder must win on focus, not breadth.
Market · Engineering orgs (roughly 100+ developers) with visible 'who-do-I-ask' friction and mature docs in Notion/Confluence; demand signal is the AI code-volume trend itself.
Pricing · Enterprise knowledge tools carry real budgets — Glean is commonly reported around $40-60/seat/month, leaving room for a cheaper focused tool.
Pros
- + Pain is growing structurally as AI code output scales.
- + Enterprise buyers already budget for this category.
- + A code-plus-docs freshness dashboard is a buildable MVP wedge.
Cons
- − Glean and other well-funded incumbents already claim expertise mapping.
- − Integration-heavy product (git, docs, chat) is a lot for a solo builder.
- − Inferring expertise from git history is noisy — last-toucher is not the expert, especially when AI wrote the code.
Existing / similar tools
Source
Hacker News (Ask HN)
The easy 80% of this product — stale-doc detection — is almost a solved problem (git blame plus doc timestamps gets you there), so the actual product is the hard 20%: inferred expertise. That is where it gets statistically noisy, because commit activity correlates with ownership, not understanding, and AI-generated commits break the correlation entirely. The realistic wedge is a dashboard for engineering leaders showing knowledge concentration and bus-factor risk per system, which is a compelling story for an engineering VP even before the expert directory works.