Concept & plan · for review

From a sentence, a finished formula.

Beauty brands tell us what they want to make. We look through a 90,000-formula library, draft a variant, and a chemist signs off before it ships.

i.At a glance

MVP build
$60–90K
Request Assistant, 6–10 wks
Full build
~$600K+
A+B+C software, phased
Run cost
$0.5–1.5M/yr
lab · chemists · reg
Corpus
~90K
in-house, never shared

ii.What this package contains

  1. The plan

    What we think the client wants, what gets built, how the data is handled, the cost, and the phased roadmap.

  2. The experience

    Thirteen panels of the brand-side journey — first-run, conversation, confirm, the no-match path, the chemist review, the revision loop.

  3. The business case

    Revenue scenarios, break-even, unit economics, cost structure, market opportunity, and how we compare to the competition.

  4. The technology

    The AI, how it finds matching formulas, and the architecture — explained for every stakeholder.

  5. Pricing & revenue

    Consumer monthly plans, the draft-and-credit model, upsells, the revenue forecast, and where the returns land.

What it is

"A high-foam, sulfate-free shampoo that smells like an ocean breeze, for the EU."

The brand types a sentence. The platform finds the closest proven formulas in the library, drafts a variant, checks it for compliance, and a qualified chemist signs off. The brand gets a production-ready spec the company stands behind.

The edge: a proprietary 90k-formula corpus + chemist-gated, production-ready output. Not a public-ingredient database. Not a workbench. Real formulas, real sign-off, real labs.

How it works · in 5 steps

  1. Describe — the brand talks to the assistant in plain language.
  2. Draft brief — the assistant structures the request; the brand confirms.
  3. Match & generate — find similar proven formulas, draft a variant.
  4. Check — automatic compliance & safety rules.
  5. Verify & deliver — a chemist signs off; the brand receives it.