01
Walletta, Inc. · Pre‑seed 2026
A cashier,
not a register.
An iPad that takes the order by voice in the guest's own language, asks the one question a good cashier asks, totals the check from the menu and sends the ticket to the kitchen.
Our customer is the owner of a counter-service restaurant. We get onto the counter as a $99-a-month self-checkout; the AI cashier inside it is the company.
Not raising: selling · our first 1,000 counters go live from sales, then a Series A on live numbers
02
Problem
The one job restaurants can't staff and guests don't want.
Cost
$120K
a year for an all-day counter: two people at about $5,000 a month each. Owners re-hire the same job every year.
Language
One
language per shift. In Los Angeles that loses the order before it starts.
Guests
1 : 3
Erewhon's newest coffee counters open with one cashier and three screens, not two cashiers and one — because guests prefer it.
Four in five restaurants in Los Angeles still have no self-service at all.
03
Product
A cashier on a stock iPad. It works today.
Speaks
42
languages, ten detected the moment the guest speaks. Every dish in the guest’s language, its menu name kept underneath.
Sells
Every ticket
It asks for the add-on and tells the owner what it earned against what it cost.
Prices
By code
The model says what the guest wants; code prices it from the menu. An item not on the menu can't reach the check.
Idle
Loop
A looped video plays until someone walks up. AI minutes bill only while it talks.
Live now: the kiosk, full-screen on any iPad. This quarter: the App Store app, Square payments, kiosk lock, offline mode, owner panel.
04
For the owner
About $170,000 a year on one counter.
$120,000counter payroll
not paid
+$52,980upsell, net of
what we charge
$172,980a year,
one counter
At 60 orders a day, the low end of a working lunch counter, and $2.70 of add-ons an order. Our side of the numbers is a slider at cashiercrew.com/model.
The screen reports its own ROI. The owner never has to take our word for it.
05
Business model
Paid per screen, and per minute it works.
$0.25
per active AI minute. We buy that minute at $0.15 today, and the cost keeps falling.
$1,595
for the full kit, paid by the owner at signing, or $99 a month to lease it.
Next
Card payments at 2.3% all-in for the owner, under Square’s 2.6% + 15¢, with 0.4% to us. Partner ads on the idle screen.
Per counter: ~$626 a month, ~$276 gross margin at 75 orders a day. ~$836 with payments.
06
Unit economics
A counter pays back in months, out of cash the merchant already paid.
To win a counter
~$1,000
10% of the counter’s first-year revenue: $100 at signing, then 10% of collected revenue for ten months. Paid only from money received.
Payback
<4 mo
on subscription and minutes alone. About two once payments route through us.
Hardware capital
$0
on the buy tier: the owner pays for the kit and the $200 install at signing.
We model 1.5% monthly churn, about 17% a year. Independents close at 15–25% a year; we install at busy counters.
07
Market
A ~$500B payroll. California first.
~$500B
a year of US pay for cashiers, counter and sales staff and their managers: about 13 million people (BLS OEWS, May 2025, our sum). That is the budget we're paid from. We sell labor, not a kiosk.
~$123M
a year in Los Angeles alone: ~12,300 counters with no self-service at ~$836 a month, payments included.
~$6.0B
a year across US restaurants: ~600,000 counters at the same price. Counts from BLS QCEW 2024 and the National Restaurant Association, 2025.
+$9B
The second product: the same screen in the aisle, no card reader, across 1.08M US retail stores at $299 a month.
1,000 counters is eight per cent of one city.
08
Competition
Everyone sells a screen. Nobody sells a cashier.
| Cashier Crew | POS · Toast, Square, Otter | Kiosk vendors | Drive-thru voice AI |
|---|
| Takes the order by talking | ✓ | — | — | ✓ |
| In the guest's own language | ✓ | — | — | — |
| Upsell on every eligible ticket | ✓, by voice | screen prompt | screen prompt | ✓ |
| Built for voice: a microphone at the counter | ✓ | — | — | ✓ |
| Runs on a stock iPad at the counter | ✓ | — | — | — |
| Takes the payment | this quarter | ✓ | ✓ | — |
Their terminals were built to be tapped, many with no microphone: a cashier means new hardware on every counter. We are voice-first, and the screen earns more than it costs.
09
Why this team wins
I have done every part of this business once.
Put unattended hardware in front of the public
Reverse-vending machines on retail floors
2016–18
Ran a counter
Opened a grocery store and worked its register myself. It became profitable.
2016–20
Built brands a stranger picks off a shelf
Twenty-one own-label brands, including Azau Crystal
2016–20
Lived the problem
An immigrant: forty taps to order one coffee
2021–25
Shipped the product
Built the live cashier myself. You can talk to it today.
2026
Engineering · Dima Sukharev and Fastshot (YC F25)
Dima: ex-Deel and Alibaba, author of OpenCommit, 7,400 GitHub stars. Fastshot: led by Dmitry Fatkhi, ex-Google AI/ML infrastructure, with engineers from Google and Cisco Meraki. Sales: 7 to 10 top closers signing now, commission-only.
Temur Valiev, founder & CEO. I closed a working business to build this one, and I will build it.
10
Where we are
Selling now. 1,000 counters from sales.
Now
Not raising. Our first 1,000 counters go live from sales, not from a round.
Proves
Owners pay every month, the add-ons out-earn our bill, and churn stays under 1.5% a month. A sold kit is paid for at signing, so kit money turns over.
Next
A Series A, raised on 1,000+ live counters and led by a fund.
Series A buys
Self-install and AI menu import, in-house engineering, a field force and the leased fleet.
Plan
125,000 counters in year five · over $1B a year · built for a million
The counters are already there, and the cashier works. We are putting it on them.
