Case study

CutMeCookie AI Product Pipeline

Moving a growing catalog onto owned infrastructure, then growing it with local AI and a person approving every design.

SunFox’s own brandLocal AIAWS Lightsail + PostgreSQL

1Problem

CutMeCookie’s cookie-cutter catalog was growing, but running it on Wix limited control over the catalog, search, checkout, and data. Manual product onboarding took about 2 hours per batch.

2Approach

We migrated the store to a self-hosted storefront on AWS Lightsail with PostgreSQL, then built an AI product pipeline with human approval:

  1. A local LLM on an ASUS GX10 analyzes catalog coverage to find under-served categories.
  2. An RTX 3090 runs image models to generate candidate designs for those categories.
  3. A person reviews and approves designs before anything ships.
  4. Approved designs get LLM-written product descriptions.
  5. New SKUs publish to the live store.

STL generation and product-image rendering are automated as well.

3Result

SunFox owns the platform and its data end to end. Product onboarding dropped from about 2 hours to under 10 minutes, and the catalog now stands at 5,100+ designs in 15,000+ styles.

All case studies

Have a similar problem?

Book a free 30-minute call with the architect who would do the work, and we'll talk through what it would take.

Book a free 30-minute call