I built a sanitized public demo of an AI customs classification assistant — looking for critical feedback

I’m Edin, and I work in customs/import-export.

I’m building an AI-powered customs classification assistant focused on HS code determination, auditability, and anti-hallucination validation.

The main project is private because it contains domain-specific customs work and beta infrastructure, so I created a sanitized public demo repo and a guest-only live demo for review.

Live demo:

The demo contains:

• no private data


• no credentials


• no Auth0


• no Convex


• no OpenAI


• no production infrastructure


• synthetic examples only


• illustrative classification outputs only

I’m specifically looking for critical feedback in these areas:

1. React / TypeScript structure


2. UI/UX clarity


3. validation flow


4. security risks


5. testing strategy


6. maintainability


7. wording that may create legal/professional overclaim risk

I do not need general praise. I need direct feedback on what is weak, unsafe, confusing, or not production-grade.

Review guide:

https://github.com/kalaba992/Carinski-ekosistem-demo/blob/main/docs/REVIEW_GUIDE.md

If you are willing to help, please open a structured issue in the demo repo using one of the templates.