The News
Buchanan Technologies, an IT managed services provider, is positioning knowledge capture as a foundational enterprise AI use case, arguing that AI systems cannot deliver reliable institutional intelligence if the underlying organizational knowledge has never been documented in the first place. The pitch comes in response to Meta’s reported development of an AI “organizational second brain” designed to convert expert corrections into reusable institutional memory. Stephen Sweett, COO of Buchanan Technologies, frames the core problem as a continuity risk: when critical operational knowledge lives exclusively with individual employees, both the business and any AI system built on top of it are operating with an incomplete picture.
Analyst Take
The framing here is simple, but the underlying problem is not. Enterprises have spent years investing in AI models, vector databases, and retrieval-augmented generation pipelines, only to discover that the quality of AI-generated answers is bounded by the quality of what was ever written down. Meta’s “organizational second brain” concept is interesting precisely because it attempts to capture expert corrections in real time, learning from the gap between what an AI says and what a human expert actually knows. That’s a meaningful architectural idea. But it sidesteps the harder question: what happens when the expert retires, leaves, or simply isn’t consulted before the AI is deployed?
The Knowledge Gap Is a Data Quality Problem in Disguise
This is where Buchanan’s argument lands with real force. Organizations across sectors carry enormous volumes of tacit knowledge, the undocumented logic behind configuration decisions, the tribal understanding of why a database schema was designed the way it was, the institutional memory of a regulatory interpretation made in 2009 that still shapes how a workflow runs today. When AI tools are layered on top of enterprise systems without first surfacing that knowledge, the outputs are not just imprecise. They are confidently imprecise, which is a different and more dangerous failure mode.
The problem is especially acute in the public sector, where ECI Research’s Google GovTech Survey found that 48.0% of respondents identified “Navigating compliance documentation and audit evidence collection” as the greatest source of cognitive load for their developers. That burden doesn’t disappear when an AI coding assistant arrives. It often intensifies because the AI inherits whatever documentation gaps already exist and then generates code or recommendations that developers must validate against standards the system was never trained to understand.
Why “Fred” Is a Risk Register Entry, Not Just an Anecdote
Sweett’s observation that every CIO should listen for the phrase “Fred is the only person who knows how that works” is not simply a good line. It describes a specific category of enterprise risk that shows up clearly in ECI Research data. According to the ECI Research Google GovTech Survey, 28.6% of respondents identified “Regression testing of undocumented business logic” as the biggest bottleneck when modernizing applications. Undocumented business logic and undocumented human expertise are two faces of the same problem. One lives in the codebase; the other lives in someone’s head.
The practical implication for ITDMs is that knowledge capture projects are not soft, cultural initiatives. They are prerequisites for AI readiness. An enterprise that deploys a generative AI assistant on top of poorly documented systems will generate faster, more confident wrong answers. The ROI case for structured knowledge capture, before AI deployment, is straightforward: it reduces hallucination risk, improves retrieval accuracy, and creates the documentation baseline that compliance and audit functions will eventually demand anyway.
For developers, the architectural consequence is just as concrete. Retrieval-augmented generation systems are only as useful as the corpus they retrieve from. If the corpus is incomplete, the retrieval layer cannot compensate.
Where This Fits in the AI Adoption Sequence
ECI Research’s Google GovTech Survey data also shows that 31.8% of respondents identified “FedRAMP/compliance approval friction for AI vendors” as the single largest blocker preventing widespread AI adoption in their developer workflows. That’s a procurement and certification problem. But the second and third most cited blockers, hallucinations and lack of trust in AI-generated code (16.7%) and data privacy and IP concerns (16.4%), are fundamentally knowledge quality problems. Trust in AI output is earned through accuracy, and accuracy depends on the completeness and integrity of the knowledge the AI draws on.
Buchanan’s positioning is smart because it targets the preparatory work that most AI vendors skip past in their rush to demonstrate product capabilities. The practical entry point they’re describing, identifying where expertise is concentrated in single individuals and systematically externalizing it, is low-risk, high-return, and genuinely foundational.
Looking Ahead
Knowledge capture will become a formal line item in enterprise AI readiness programs over the next 12 to 24 months. As organizations move from AI pilots to production deployments, the failure patterns will increasingly point back to incomplete organizational knowledge rather than model capability gaps. Vendors offering structured knowledge elicitation, expert interview frameworks, and documentation tooling integrated with AI pipelines will find a real market, particularly in regulated industries where the cost of a confident wrong answer is measurable.
Buchanan Technologies is making a credible early move by naming this problem clearly and offering advisory framing around it. The competitive question is whether they can build a repeatable delivery methodology fast enough to own the space before larger systems integrators and hyperscalers absorb knowledge capture into their broader AI transformation offerings. The window is real, but it’s not indefinite.
Stay Ahead of Application Development Trends
Get weekly analyst insights, research notes, event coverage, and AppDevANGLE updates delivered directly to your inbox.
Subscribe for Weekly Insights
Join technology leaders, practitioners, and GTM teams following the trends shaping modern software delivery.
Looking for deeper research access?
Explore ECI Research reports, survey insights, and market analysis through the ECI Research Portal.
