Translate the opportunity into a product one-liner.
The product brief defines the user, promise, scope, non-goals, constraints, and acceptance criteria.
Back to Playbook
Operating model
Helm
Set enough direction that the work can be built without drifting from the product intent.
01Context
Direction before execution.
Helm is the steering layer. It turns a promising idea into enough context that a human builder and AI collaborators can work without constantly inventing the product as they go. This is where vague ambition becomes a product dossier.
The value of Helm is not documentation for its own sake. The docs become the working context: product brief, user stories, flows, architecture, constraints, security and privacy boundaries, metrics, AI roles, and the review expectations that keep the work inside the product intent.
02Model
The context system before execution.
The model is the part of the article where the idea becomes usable: a sequence of decisions, artifacts, or checks that can guide real product work.
Turn intent into workflows.
User stories and UX flows make the product navigable for humans and legible to AI collaborators.
Name the technical and trust boundaries.
Architecture, domain model, APIs, permissions, security, privacy, cost, and performance become build constraints, not surprises.
The model moves from question to artifact to evidence.
- BriefTranslate the opportunity into a product one-liner.
- FlowTurn intent into workflows.
- SystemName the technical and trust boundaries.
03AI collaboration
Helm is where prompting becomes product direction.
The Playbook treats AI collaborators like a small product team. Each role needs a job, context, constraints, expected output, and human review boundary.
That context normally includes a current-state assessment, source-authority index, product principles, domain and workflow model, decision log, risk register, execution plan, validation matrix, launch checklist, closeout, weekly product note, and experiment log. These are not ceremonial docs; they are the rails that keep AI work aligned.
This is why Helm matters for product roles in the AI era: it proves that AI speed can be governed by product judgment.
04Builder practice
The dossier is not bureaucracy.
A good dossier reduces rework. It helps the implementer avoid inventing product behavior, helps the reviewer judge the result, and helps the builder remember why choices were made.
Helm also defines human decision rights. The human owns purpose, positioning, target users, taste, scope, sequence, legal/privacy/security/financial risk acceptance, customer communication, deployment, public claims, and decisions to expand, pause, revert, or remove work.
The practical test is whether an AI collaborator can load the relevant context, understand the role, produce the requested artifact, surface assumptions, and stop at the right boundary without silently changing the product.
AI can recommend, implement, and verify. It should not silently decide what the product is.
From this essay05Artifacts
What this leaves behind.
I use artifacts as evidence of thinking. They make product judgment reviewable, reusable, and easier to connect back to the work.
Product brief
The source truth for what the product is, who it serves, what it does first, and what it will not do yet.
Personas and stories
Roles, jobs, acceptance criteria, and priority paths for the first usable product surface.
UX flows
The sequence of screens, states, empty/error conditions, and handoffs that define the experience.
Architecture and constraints
The system map, data model, integrations, non-functional constraints, and risk boundaries.