Team coherence at scale.
WaitlistYour team is using AI. Everyone has their own approach. The brand designer prompts one way, the copywriter another, the junior designer a third. The output is inconsistent. Voice drifts between contributors. Nobody's evaluation criteria match. You're the one who has to catch it all, and you're catching it too late.
Individual AI practices don't scale. A single practitioner can hold voice, evaluation, and context in their head. A team of five can't. Each person's AI workflow produces output that reflects their prompting habits, not the project's standards. The CD role shifts from directing creative work to policing AI output, and that shift happens quietly. You don't notice until the deliverables feel off and you can't explain why.
The Accommodation Design methodology gives you the individual practices. This track makes them collective. Shared voice protocols replace individual prompting styles. Shared lens arrays replace individual evaluation instincts. Shared savepoint infrastructure replaces individual context management. The CD directs instead of policing because the system holds what the team needs to hold.
Case-based: building a team voice protocol that holds across contributors
A style guide tells people what to do after they've written something. A voice protocol constrains what gets generated in the first place. The difference matters when five people are using AI to produce copy, concepts, and deliverables for the same brand. This module teaches you to build a team voice protocol: the shared constraint document that governs AI output across every contributor on the project.
You start with the house style as it actually exists, not the brand guidelines PDF that nobody reads. You extract the real voice from real output that worked, build rules from those patterns, and test the protocol with multiple contributors generating for the same brief. The protocol becomes a generation constraint, not a style guide. Everyone's AI output sounds like the brand because the constraint runs during generation, not during review.
Deliverable: A team voice protocol tested across multiple contributors on a shared project. House style as generation constraint, not post-hoc correction.
Applied leadership: shared lens arrays and team-level evaluation standards
When one person evaluates their own AI output, they use their own standards. When five people evaluate their AI output, you get five sets of standards. The CD becomes the reconciliation layer, reviewing everything against criteria that exist only in their head. This module externalizes those criteria into shared lens arrays that every contributor can run independently.
You build shared lenses: brand voice lens, technical accuracy lens, audience appropriateness lens. Each contributor runs the same evaluation stack on their own output before it reaches the CD. Savepoints become shared infrastructure across Discord, Slack, and project channels. Multiple contributors mark what matters in the same project using the same protocol. The CD reviews evaluated work, not raw output.
Deliverable: A shared lens array and savepoint practice deployed across a real team or project channel. Evaluation standards that run at the contributor level.
Applied leadership: the coordinator that the whole team runs from
The individual coordinator wires one person's practices into a pipeline. The team coordinator wires a group's practices into a shared system. This module teaches you to use contributors as decomposed agents: each person running a specific part of the methodology, feeding results to the coordinator that holds the whole project together.
You build collective research and decision-making workflows using the same synthesis steps from the methodology, applied across multiple people. The coordinator becomes the document the whole team runs from: who handles input, who runs evaluation, how decisions flow, where savepoints live. The CD role returns to direction because the system handles coherence.
Deliverable: A team-level AI governance system. Shared voice protocol, shared lens array, shared savepoint practice, team coordinator. The system holds coherence across contributors so the CD can direct instead of police.
The team coherence problem is a version of the same problem that appears in every domain this methodology touches. A 13-year enterprise design system serving multiple teams, each building independently, needed shared governance to hold brand coherence across contributors. A household with four people and different needs required shared protocols so the system worked for everyone, not just the person who designed it. The practices scale the same way: externalize the standards, constrain during production, evaluate against shared criteria.
The voice protocol infrastructure used across all Joinery operations was built for exactly this problem: multiple agents and contributors producing output that needs to sound like one coherent voice. Forty rules, tested in production, governing generation across every output channel.
This track is a future offering. Pricing and structure to be announced. See all offerings.
Other specialist tracks: Brand Designer · Copywriter · Fiction Writer · Nonfiction Writer · Educator
Future offering. Join the list for updates.