Procurement matching · Guided decision flow
Yunnan Coffee Match
A real six-step procurement-matching prototype for turning a business scenario into a clearer coffee product direction and buyer brief.

What this project is
A real build with its status left visible.
An independent industry prototype that combines a narrative Yunnan coffee presentation with a six-step matching flow and a structured procurement summary.
The formal website presents verified case-study evidence; the real local project depends on server routes and local inquiry storage.
Cafes, hospitality operators, wholesalers, and food brands.
Problem / need
The specific friction behind the interface.
A buyer may know the operating scenario but not the right coffee format, purchasing scale, or questions to confirm. The project tests how a guided flow can structure that ambiguity before human follow-up.
What was built
The working parts, not an inflated result story.
- 01
A narrative Yunnan coffee industry homepage
- 02
A six-step procurement-matching flow
- 03
Rule-based primary and alternative recommendations
- 04
A structured procurement summary with clear confirmation boundaries
Experience structure
How someone moves through the project.
- 01
Begin with the operating context rather than a generic catalogue.
- 02
Collect product direction, purchasing scale, priorities, packaging needs, and timing.
- 03
Generate a rule-based recommendation and alternative direction.
- 04
Summarise the buyer brief while keeping price, inventory, and delivery subject to human confirmation.
Key screens
Interfaces captured from the verified project runtime.




Deliverables
- Narrative product-origin page
- Guided procurement matching flow
- Rule-based recommendation screen
- Buyer brief summary
Functional scope
- Six-step guided questionnaire
- Rule-based recommendation logic
- Alternative product direction
- Procurement summary
- Privacy and data-handling disclosure
Technical stack
- Next.js
- React
- TypeScript
- Motion
- SQLite
Verified from the clean feature/demo-data-privacy-closure-v1 source branch.
Evidence
- Canonical source root is recorded in the local project audit.
- Build, lint, typecheck, six-step flow, recommendation, and procurement summary were verified in the recovery sprint.
- Real screenshots are stored under public/projects/yunnan-coffee-match/real/.
Delivery relevance
Where this kind of project can be useful.
The same pattern can support cafes, hospitality groups, wholesalers, specialist retailers, and other businesses that need a guided product-selection or qualification flow.
Can be adapted into a guided selection, quote intake, or product-fit tool for specialty suppliers and service businesses.
Case-study-led workflow design with a scoped frontend and server boundary when submission or storage is required.
Who it is for
- Cafes and hospitality operators
- Food and beverage brands
- Wholesale buyers
Current boundary
What this evidence does — and does not — claim.
- The content is an industry workflow demonstration, not a claim about a specific coffee company.
- Recommendations do not promise price, stock, minimum order, or delivery.
- The formal website does not copy private environment files or local inquiry data.
- No public same-origin demo is exposed because the real project has API and local storage boundaries.
- Recommendations require human confirmation before any commercial use.
Next step