The methodology exists. It's in a Google Doc from 2021, a partner's slide deck, and someone's head.

The situation

Institutional knowledge that only works if you already know where to look

The firm runs three practice areas — operations, digital transformation, and organizational design. Each practice has its own frameworks, but they evolved independently. Operations has a "diagnostic sprint" method documented in a 47-page Google Doc that was last updated eighteen months ago. Digital transformation has a delivery playbook that lives as a slide deck a senior manager presents during onboarding. Organizational design has a set of interview protocols that exist only as templates in one partner's Drive folder, shared on request.

Proposal writing is where the fragmentation hurts most. A consultant drafting a proposal for an operations engagement searches the shared Drive for "diagnostic sprint template," finds four versions from different years, and has no way to know which one reflects current methodology. They end up asking a partner, who sends them a link to a fifth version they didn't know about. The proposal goes out with a framework description that is subtly different from what the delivery team actually does — because the proposal template and the delivery playbook were never reconciled.

The firm recently started using an AI assistant to help draft proposal sections and prepare client briefings. The assistant has no access to any of this knowledge because it is scattered across personal Drives, locked in slide decks, or never written down at all. Every AI-drafted section is generic because the firm's specific methods, terminology, and differentiators are not in any format the assistant can ingest.

The work

What Taz actually does with this knowledge

Taz inventories every document the firm identifies as a method, framework, template, playbook, or reference — across Drives, email attachments, and whatever partners point to. The inventory is exhaustive within the scope the firm grants access to.

• Knowledge inventory: Every document is cataloged with its title, location, last-modified date, apparent owner, practice area, and content type (framework, template, playbook, glossary, pricing model). Duplicates and version conflicts are identified. Documents that reference each other but contradict each other are flagged. The inventory answers "what do we actually have, where is it, and how stale is it?"
• Content extraction and structuring: Key documents are extracted from their native formats (Google Docs, Slides, Word, PDF) into markdown with YAML frontmatter. Each bundle includes metadata — title, owner, practice area, last-reviewed date, status (current/draft/superseded), and related documents. Slide decks become text-and-diagram documents. Oral knowledge captured during partner interviews becomes documented procedures with attribution.
• Cross-linking and hierarchy: Documents are linked into a navigable structure. The operations diagnostic sprint references the standard interview protocol, which references the findings template, which references the proposal language for that engagement type. A consultant or AI assistant starting from any entry point can follow the chain to everything related.
• Gap identification: The inventory reveals what is missing — a practice area with frameworks but no delivery playbook, a proposal template that references a pricing model nobody can find, an onboarding process that assumes knowledge never written down. Each gap is documented with a recommended owner and priority.
• Index and search layer: A master index maps every knowledge asset by practice area, document type, and engagement phase. Structured for both human browsing and AI ingestion — an assistant asked to "draft the discovery section for an ops engagement" can locate the correct framework, current template, and example language from the index.

The deliverable

What the firm receives

• Knowledge inventory spreadsheet — Every identified document: title, location, owner, practice area, content type, last-modified date, status, version conflicts, and a completeness score. The firm can see its entire knowledge estate in one view.
• OKF knowledge bundles — Structured markdown files with YAML frontmatter for each method, framework, template, and playbook. Portable across tools — works in Notion, Confluence, a git repo, or any AI assistant's context window. Cross-linked by practice area and engagement phase.
• Master index — A navigable map of all knowledge assets organized by practice area, engagement phase (discovery, proposal, delivery, close-out), and document type. Includes the recommended "start here" path for each common task.
• Gap register — Missing documents, unresolved version conflicts, and undocumented procedures. Each entry has a recommended owner and a priority ranking based on how often the gap causes rework or delays.
• Maintenance checklist — A quarterly review protocol: which documents to check, who reviews them, what triggers an update, and how to add new knowledge assets to the bundle without breaking the structure.