Consulting firms lose money when project knowledge stays trapped in finished engagements. Dock fixes the reuse problem by giving the agent a place to draft case-study updates from recent work and routing each draft to a named senior consultant for approval. The agent does the reading. The senior reviews and attributes. The firm gets a current, citable knowledge base instead of a dead wiki.
Notion, Confluence, and Google Workspace stay the system of record for the raw data. Dock is the system of record for what the AGENT INTERPRETS. Each Dock row carries a pointer back to the platform record, agent identity, decision, reviewer, and timestamp. The agent re-fetches platform data via fresh API reads when it needs current state.
The Dock surface: Case-Study Drafts
| Engagement | Source pages | Agent draft summary | Drafted by | Senior reviewer | Status |
|---|---|---|---|---|---|
| Acme retail margin rebuild | Notion: 14 pages, Confluence: 3 retros | Repricing model lifted gross margin 4.2 pts over 11 weeks; key lever was SKU-level elasticity scoring | agent:dock-knowledge-bot | sarah.chen@firm.com | Approved 2026-05-22 |
| Northwind ERP cutover | Drive: 22 docs, Confluence: cutover runbook | Phased cutover across 6 regions; vendor reconciliation pattern is reusable | agent:dock-knowledge-bot | priya.iyer@firm.com | Changes requested |
| Helio Health intake redesign | Notion: 9 pages, Drive: 3 interview transcripts | Patient-intake redesign cut average admit time 18 minutes; methodology section needs PHI scrub | agent:dock-knowledge-bot | (unassigned) | Awaiting reviewer |
Each row links back to the underlying Notion page IDs, Confluence space keys, and Drive file IDs. The agent does not paste content. It writes an interpretation and stores the pointer.
The workflow
A partner closes an engagement and tags it complete in Notion. The Dock knowledge agent picks up the signal, reads the engagement workspace through the Notion API, pulls the matching Confluence retros, and skims any Drive transcripts. It drafts a case-study update following the firm's template and writes a new row to the Case-Study Drafts table. The row names the agent, lists every source document, and assigns a senior consultant who worked the engagement. The senior opens the draft in Dock, sees the agent's interpretation next to live links to the source platforms, and either approves, edits, or requests changes. Approved drafts publish back to the firm's Confluence knowledge base with a footer that names both the drafting agent and the approving consultant. For a deeper read on why attribution matters, see Drafted by agent.
This loop matches what KCS Academy calls capture-in-workflow: knowledge documented as a byproduct of solving the problem, not as a separate after-hours task (Service Innovation Consortium).
Why it matters
Source Global Research's 2026 work on thought-leadership quality flagged a "sea of sameness" problem across consulting firms (Source Global Research). Firms publish, but the publications do not draw on the firm's own recent project evidence. Dock closes that gap by routing recent engagement data into reviewed case-study updates that carry both agent and senior-consultant fingerprints. Reviewers see exactly what the agent read and exactly what it concluded. The firm gets a knowledge base that is current, attributed, and safe to cite in proposals.
The same architecture supports adjacent practices. The research practice uses it for literature reviews. The design practice uses it for pattern libraries. The IT practice uses it for an internal knowledge base keyed to ticket trends.
Pair this with the broader Dock for consulting and agencies pillar for the full operating picture.
Start a Dock workspace for your knowledge practice
FAQ
Does the agent edit the original Notion or Confluence pages? No. The agent reads through the API and writes interpretations into Dock rows. The source platforms stay untouched until a senior consultant approves a draft and chooses to publish it back to Confluence.
How does a reviewer know what the agent actually read? Every Case-Study Drafts row lists the exact Notion page IDs, Confluence space keys, and Drive file IDs the agent consumed, with timestamps. The reviewer can open any source in one click and compare it to the draft.
Who gets credit on the published case study? Both the drafting agent and the approving senior consultant. The footer reads "Drafted by agent:dock-knowledge-bot, approved by [consultant name]" with the approval timestamp. See agent identity for the underlying model.
What stops the agent from leaking client-confidential material? The agent operates under its own identity with scoped Notion, Confluence, and Drive permissions. It cannot read workspaces it is not invited to, and every read is logged. Senior reviewers see the source list before approving, so PHI or PII issues surface before publication.