Claude Code builds a live coffee-shop order workflow
timhartmann7 reports building and selling a coffee-shop system with QR table ordering, a barista CRM, staff logins and Telegram loyalty, now serving real customers.
reddit.com
Suggested prompt
Suggested prompt written by Agents In Use, not the source author. Replace bracketed inputs and review tool permissions first.
Build a reviewable coffee-shop ordering workflow where table QR codes preserve the table number, staff can submit orders to a barista queue, and Telegram customers receive the agreed loyalty status. Show the owner the actual order path before rollout.
Reach people exploring AI agents
See placements →Evidence at a glance
Sourced account- Source author
- timhartmann7
- Original account
- Read the source ↗
- Published
- Source link checked
- Reported result
- timhartmann7 reports building and selling a coffee-shop system with QR table ordering, a barista CRM, staff logins and Telegram loyalty, now serving real customers.
- Human involvement
- The developer handles sales, owner feedback, deployment and maintenance. Production stability and the sale price are the author’s reports.
A linked account supports attribution. It does not mean we independently reproduced its result.
Report an issue with this case
Open correction emailOpens your email app with the case link included. Nothing is sent automatically.
How the agent did it
- Build a branded table-order demonstration with Claude Code and SvelteKit.
- Show a working interface to the shop owner and agree requirements.
- Connect QR table numbers to the barista order queue and staff logins.
timhartmann7 uses Claude Code to build an operational ordering system for a local coffee shop. An initial branded SvelteKit demonstration helps the developer show owners a working interface; the first buyer agrees to pay seven hundred dollars according to the account.
Guests scan table-specific QR codes and place orders on their phones. A barista CRM receives the items and table number. Separate staff logins let waiters submit several tables’ orders without walking back to the register for each one. Telegram login supports a loyalty programme: after five orders, the customer retains a twenty-percent discount. The author also ships a Telegram mini-app version.
The core build takes three days, followed by about a week of fixes and owner sign-offs. The author reports the system serving customers in production and returning monitoring logs. These are personal claims, without an independently audited reliability or sales comparison.
Claude Code is the agent that constructs the workflow; the finished ordering app is ordinary software, not an autonomous restaurant agent. The human developer finds the buyer, interprets the owner’s needs, deploys the app and maintains it. The useful pattern is to demonstrate the actual table-to-bar order path and loyalty state, then obtain operational feedback before treating an attractive demo as a finished shop system.
Prompt
Suggested prompt (written by Agents In Use, not the author's)
Source: Reddit · timhartmann7
Frequently asked questions
Which part is the agent?
Claude Code performs the engineering. The author explicitly says the sold ordering application is not itself an agent.
How does the bar know the table?
Each QR code carries its table number, which appears with the ordered items in the barista CRM.
Was the first demo the finished product?
No. The author reports three days of core work and another week of fixes and owner sign-offs before the production account.
More details
Remaining steps
- Add Telegram login and the agreed five-order loyalty rule.
- Complete owner sign-offs and inspect production monitoring after deployment.
About this account
Claude Code constructs the ordering and loyalty software from the developer’s instructions; the deployed application processes normal customer and staff actions.
Firsthand account by timhartmann7; original publication date verified against Reddit JSON created_utc 1781801774 (2026-06-18 UTC). Outcomes are the author’s reports, not independently reproduced. Suggested prompt written by Agents In Use; not the author’s original prompt.
Credit: timhartmann7
Source published
Added to Agents In Use
Source checked
Tools and topics
Claude Code · SvelteKit · QR table codes · Barista CRM · Telegram mini app
Coffee-shop table orders · Staff order queue · Five-order loyalty
Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.