---
title: "Dock + Lattice: performance-review prep with agent-drafted briefs"
excerpt: "Pair Lattice as the system of record for review cycles with Dock as the workspace where the agent drafts manager prep briefs, captures sign-off, and leaves a per-row audit trail."
author: mei
category: Use Cases
date: "2026-05-30"
---

Managers walk into review conversations underprepared because the source material lives in five places: Lattice goal updates, Culture Amp engagement signals, 1:1 notes, project trackers, and last cycle's calibration. Dock pairs with Lattice to fix the prep problem without replacing it. Lattice still owns the review cycle and the final written record. Dock holds the agent's drafted brief for each direct report, the manager's edits, and a reviewer sign-off before anything posts back. The result is a manager who walks in ready and a cycle the People team can audit row by row.

Lattice and Culture Amp stay the system of record for the raw data: review responses, goal status, eNPS, and engagement scores. Dock is the system of record for what the agent interprets from that data. Each Dock row carries a pointer back to the platform record through a `lattice_review_id` or `cultureamp_survey_id` field, plus the agent's identity, the draft it produced, the reviewer who approved, and the timestamp. When the manager opens the brief, the agent re-fetches the underlying Lattice and Culture Amp records via fresh API reads so nothing goes stale between draft and conversation. See [Dock for People Ops](/blog/dock-for-people-ops) for the broader pattern and [Dock for HR](/blog/dock-for-hr) for adjacent workflows.

## The manager prep brief table

| review_id | report | lattice_review_id | agent_draft_status | manager_edited | reviewer | approved_at |
|---|---|---|---|---|---|---|
| PREP-204 | A. Okafor | LAT-9921 | drafted | yes | hr-partner-jess | 2026-05-28 14:02 |
| PREP-205 | M. Liu | LAT-9934 | drafted | no | hr-partner-jess | 2026-05-28 16:40 |
| PREP-206 | R. Patel | LAT-9941 | needs_more_data | pending | hr-partner-jess | null |

## One worked workflow

The review-prep agent runs on a cycle trigger from Lattice. For each direct report, it reads goal progress, peer feedback summaries, and the prior-cycle calibration notes, then pulls the matching Culture Amp engagement record. It drafts a one-page brief into a Dock row: themes to raise, specific examples with links back to Lattice, suggested development areas, and three discussion prompts. The row enters `drafted` status. The HR business partner reviews, edits anything thin, and marks `approved`. Only then does the brief unlock for the manager, and only the manager can post the final review back to Lattice. The agent never writes to Lattice directly. See [agent audit and compliance](/blog/agent-audit-and-compliance) for the reviewer pattern and [agent identity](/blog/agent-identity) for why the draft is attributed to a named agent, not a service account.

## Why it matters

Review conversations get better when the manager has read the material. SHRM's performance management guidance frames thoughtful preparation as the lever that turns reviews from compliance theater into useful conversations ([SHRM, Managing Employee Performance](https://www.shrm.org/topics-tools/tools/toolkits/managing-employee-performance)). Josh Bersin's research on the manager gap argues the same point from a different angle: organizations underinvest in the act of preparing managers to manage ([Josh Bersin, The Manager Gap](https://joshbersin.com/)).

Dock does not replace Lattice's review workflow or Culture Amp's measurement instrument. It captures the interpretive layer that today lives in a manager's head or a stray Google Doc. Every brief has a named drafter, a named approver, and a fresh read of the source record. When People Ops asks how a calibration decision was reached, the answer is a row, not a memory.

The audit trail also matters for the People team that owns the cycle. Calibration sessions reference the approved briefs. Promotion committees cite them. Six months later, the same agent can pull a row by `lattice_review_id` and surface what was actually said the last time around. See [Dock for HR](/blog/dock-for-hr) and [agent identity](/blog/agent-identity) for the identity model that makes this trail trustworthy.

## Try it

Bring your Lattice cycle into a Dock workspace and let the prep agent draft briefs for your next round of reviews.

## FAQ

**Does Dock replace Lattice or Culture Amp?**
No. Lattice owns the review record. Culture Amp owns the engagement instrument. Dock holds the agent's interpretation and the reviewer's sign-off.

**Who can approve a brief before the manager sees it?**
The HR business partner assigned to that org. The reviewer field is required, and a brief cannot move from `drafted` to `approved` without a named human.

**What happens if Lattice data changes after the brief is drafted?**
The agent re-fetches Lattice and Culture Amp records when the manager opens the brief. Stale drafts are flagged for re-approval before the conversation.

**Can the agent post the final review back to Lattice?**
No. The manager writes the final review in Lattice. The agent's authority ends at the approved Dock brief.