---
title: "Dock for founders: board-deck prep workflow with attributed co-founder review"
excerpt: "Dock turns board-deck prep into an attributed workflow: the agent reads metrics from Notion, qualitative inputs from Google Workspace, drafts slides in Pitch, and routes the narrative to co-founders for line-level review."
author: mei
category: Use Cases
date: "2026-05-30"
---

Board-deck prep is the worst week of every quarter for most founders. The deck lives in Pitch, metrics in Notion, qualitative inputs across Google Docs and Sheets. By the time the narrative is stitched together, no one remembers which number came from where or which co-founder agreed to which framing. [Dock for founders](/blog/dock-for-founders) fixes that: the agent drafts the deck and records the review trail row by row.

Pitch, Notion, 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 board-narrative surface

One Dock table per board meeting. Each row is one slide or claim.

| Slide | Claim drafted by agent | Source pointers | Reviewer | Status |
|---|---|---|---|---|
| 03 Revenue | "Net new ARR was 412k, up 38 percent QoQ, driven by mid-market" | notion://metrics/q2-arr, gsheet://finance/cohorts | jordan@ (CFO co-founder) | approved |
| 07 Hiring | "Closed 4 senior eng hires; CTO bandwidth now the constraint" | gdoc://people/hiring-log, gdoc://exec/ctoraq | priya@ (CTO co-founder) | edited-then-approved |
| 11 Risks | "Top risk: one customer is 22 percent of ARR" | notion://customers/concentration | jordan@ | pending |

Each row links to the Pitch slide and the platform record. Nothing is retyped. The drafting agent identity sits on the row, separate from the human reviewer, the way [agent identity](/blog/agent-identity) requires.

## The workflow

1. Board date is set. The agent (`deckwriter-v2`) opens the Dock table and pulls last quarter's slide skeleton.
2. It re-fetches: ARR cohorts from Notion, runway from a Google Sheet, hiring log from a Google Doc.
3. For each slide it writes one claim row. Source pointers are required; the agent cannot save a claim without them.
4. Co-founders get a review queue filtered to their rows. They approve, edit, or kick back. Edits stay on the row.
5. Once every row is approved, the agent renders the Pitch deck. Reviewer initials land in the speaker-notes footer.

This is the [agent collaboration primer](/blog/agent-collaboration-primer) shape: the artifact (the deck) is downstream of attributed claims, not the other way around.

## Why it matters

Board members read decks for signal about how founders think. When the CFO co-founder edits the revenue claim and the CTO co-founder edits the hiring claim, the deck stops being "the agent's draft" and becomes a real co-founder narrative, with the trail to prove it. NVCA model documents assume directors receive materials prepared with care; an attributed trail is the cleanest evidence of that care.[^1] The Harvard Law School Forum on Corporate Governance has tracked rising board-material scrutiny across 2025-26.[^2]

This is [Cloud 2.0 for product](/blog/cloud-2-0-for-product) applied to founder ops. The same pattern carries to investor briefings: see [Dock for investors](/blog/dock-for-investors).

## Single CTA

Want this board-prep table in your workspace before next meeting? Copy the board-narrative template from the Dock for founders pillar.

## FAQ

**Q: Does Dock store the deck itself?**
A: No. Pitch stores the deck. Dock stores the attributed claims, with pointers to Pitch slide IDs and the Notion or Google Workspace records they came from.

**Q: What if a metric changes after a row is approved?**
A: The agent re-fetches before render and flags any approved row whose number has moved past a threshold. The row reverts to pending; the co-founder sees old framing and new number side by side.

**Q: How is this different from a service account writing to Pitch?**
A: A service account leaves "Pitch API" as the actor. Dock requires the agent to act under its own identity (`deckwriter-v2`), separate from the reviewer. That separation makes the audit usable. The argument lives in [agent identity](/blog/agent-identity).

**Q: Can the agent draft talking points too?**
A: Yes. Talking-points rows hang off the same slide row, with the same reviewer field, and render into Pitch speaker notes.

[^1]: National Venture Capital Association, Model Legal Documents (October 2025 revision). https://nvca.org/model-legal-documents/
[^2]: Harvard Law School Forum on Corporate Governance, ongoing coverage of board oversight and disclosure practices. https://corpgov.law.harvard.edu/