Outgrew the reader-first till.
A till built around the kitchen, the inventory and the audit — not bolted onto a card reader after the fact. The till that already knows the rules.
What a modern floor OS does in 2026.
This is the objective standard — not Vetka's claim, the baseline a serious operator should expect from any till sold today. Read it as a checklist. Tick what your current system already does.
Runs on phones your team already owns
No fleet of proprietary tablets to buy, charge and replace.
Cloud signing on every receipt
No fiscal box per device. Compliance signed in the cloud, country-correct.
Snap a menu, AI does the data entry
Photo → categories, prices, modifiers, allergens. Five minutes, not five hours.
Recipes deduct stock as you sell
Variance flagged at the daily threshold. Theft visible before payroll closes.
Tips routed by hours, section or role
Logged as a separate ledger line. Tax-clean, payroll-clean, dispute-clean.
Card + cash + tap-to-pay in one screen
No mode switching mid-queue. Apple Pay and Google Pay first-class.
Day close generated automatically
Hash-chained, signed, retained. Auditor opens in a tab.
Multi-location ledger when you need it
One product, one bill, one portfolio view. Add a venue, not a vendor.
Tick what your current till does.
Your score
0/8
Vetka does all 8. Your gap is the case for switching — read it back to yourself before you talk to us.
Problems we hear from operators ready to switch.
Generic failure modes from owners who outgrew their first POS. We don't name the system — if any of these match your Tuesday, you're the reader we wrote this for.
Card reader was the product, the till came later
A reader-first stack treats the POS as a follow-up. Floor flow, kitchen display, recipes, day-close — all added on top of an app that started as a payments app. A floor OS works the other way around: the floor is the centre, payments slot in.
Card-fee surprises on the month-end statement
Headline rate, processing rate, currency conversion, statement fee, withdrawal fee — the real take per transaction is impossible to compute in advance. Bundled, transparent card share is the bar in 2026.
No real recipes, no real variance
Counter sales worked. The first time the kitchen lost three bottles between stocktakes the system had nothing to say. Recipes that deduct stock and a variance threshold are the difference between knowing and guessing.
Compliance as a sticker, not a system
A till that signs the receipt is not the same as a till that runs the audit. Hash-chain, retention, auditor-ready export, country-correct format — these are the system, not the sticker.
Same job, every day. Different cost.
How the same recurring jobs run on each system. Factual. Yours to verify.
- Take a cardReader, separate processor relationshipBundled into tier — 0.5% to 2.5% all-inStatement reading
- See real costProcessing + statement + conversion feesOne number on the pricing pageFee variance
- Recipes deduct stockManual count, separate inventory appSale fires, recipe deducts, variance flagsInventory app cost
- Kitchen displayAdd-on or printer-onlyKDS in the box, course-awareKDS subscription
- AI menuManual entry, one item at a timePhoto of menu → categorised, priced, modifier-mapped~4 hr / menu
- Daily closeSales report, no signed closeAuto at 23:59. Signed, archived, sent~30 min / day
- AuditPull from app, build packetOpen in a tab. Hash-chained, retainedHalf a day
We don't pull customers.
We're here when you outgrow.
If your current till runs the floor, the kitchen and the books without your team noticing it's there — stay. We don't cold-call, we don't cut commercial deals to flip you. The day your old system costs you a Saturday night, this page is still here.