Brieflo is currently available by invitation only.

How it works

Who does what

The customer describes the problem and approves functional and visual choices. The developer sets price, technical direction and planning, and builds or delegates the work.

1. Interview

Brieflo asks about business goals, audience, users, the current situation, problems, processes, desired features, roles, content, branding, integrations, planning, success criteria and what is deliberately out of scope. Answers are stored immediately before any further AI step, so a network or provider failure never discards an answer.

2. Agreements and approval

Every requirement has a stable ID, source references, acceptance criteria and an epistemic status: confirmed, inferred, ambiguous, contradictory or not discussed. An AI proposal never becomes a confirmed customer wish on its own.

3. Quote or no quote

After discovery you choose whether to produce a quote. If you skip it, that choice stays visible and Brieflo invents neither a price nor a customer acceptance.

4. Brand, theme and screens

Logos, design directions and screens are version-bound assets with approvals. When something changes, dependent items are visibly marked as possibly stale.

5. Technical plan and handoff

Technical decisions carry a dependency graph. Change the hosting choice and Brieflo shows which other decisions must be re-evaluated. The agent prompt and ZIP are only finalised once the readiness check finds no hard blockers.

6. Development and feedback

You record releases and deployments. Your customer can then submit bugs, small changes and new wishes. Those follow an explicit status flow, so a rejected request never flows into development on its own.

{# The consent choice is reported to the server, so the CSRF cookie must be present on every public page. Without it the report is refused and the decision stays local-only. #}