The restaurant OS that reorders before you run out.
Every core POS component — plus a fleet of AI agents that catches what you'd miss and hands you the fix. You just approve.
Real, persisting slices — not a mockup Free tier for dhabas
Shift Commander · Tadka House
AI · liveBefore you open: 2 items below the safe line, with a +105% Wed spike incoming. 4 actions queued.
Reorder Paneer to par
Critical · 2 kg on hand
Set Chicken 65 delivery to ₹255
Losing ₹2/order on delivery
Prep for a +105% spike · Wed
Ganesh Chaturthi eve
Win back Rohan — last order 1 wk ago
Silver · ₹6,100 lifetime
Grounded · approval-gated · every action logged
One dinner rush
What a shift looks like with the fleet on.
The rush is coming.
Rai saw tonight's +105% spike days ago. Prep and staffing, already sorted.
Ganesh Chaturthi eve — sweets & thali demand up ~80%.
Paneer's about to run out.
Reorder to par — drafted for your supplier. You just tap approve.
A dish is quietly losing money.
Delivery eats the margin. One tap sets a price that clears the cut.
Chicken 65 · per channel
Table 4 is waiting too long.
Caught before it becomes a one-star review — expedite, now.
Table 4
22 min · food not served
⏱ expedite now
The drawer matches.
Reconciled to the rupee. Every approved action, logged.
Ganesh Chaturthi eve — sweets & thali demand up ~80%.
₹9.4B → $17.9B
Global restaurant POS market, 2024 → 2032 projection.
Of India's organised POS market held by three incumbents — none forecasting-first.
Of India's cloud kitchens are independent 1–3 brand operators — the segment enterprise POS ignores.
The agent fleet
Thirteen specialists. One Commander. Zero silent changes.
Each agent watches one slice of the restaurant and proposes with the exact facts. The Shift Commander ranks them; you approve. Nothing touches your menu, prices, orders or messages without a tap — and every action is logged.
Shift Commander
OrchestratorRanks every agent's calls into one approve-list — the orchestrator.
Service Recovery
Late tickets and unclosed bills, live — protects reviews during the rush.
Stock & Procurement
Reorder to par, stock-out 86s, supplier reorder drafts.
Menu Margin Guardian
Flags dishes losing money by channel after commission & packaging.
Demand & Prep
Turns forecast spikes — rain, IPL, festivals — into prep & staffing.
Cash Close
Unreconciled money and open bills, caught before the day locks.
Guest Recovery
Wins back lapsed regulars with opt-in, approval-required messages.
Chain Control
Escalates only material cross-outlet variance — not noise.
Procurement
Flags ingredients you overpay for vs the network, and open pooled buys.
Menu Engineer
Spots the Dogs — low-margin, low-selling dishes to rework or retire.
Reputation
Drafts a reply to every negative review — you approve before it posts.
Loss Prevention
Catches outlets running hot on voids, discounts and comps.
Ad Performance
Cuts aggregator campaigns bleeding below break-even ROAS.
Payout Recovery
Reconciles aggregator payouts and drafts the claim for every mis-settled order.
Beyond a POS
The things only Rai does
Run the restaurant by chatting, let agents act within limits, and tap capabilities no incumbent offers — each one live in the demo.
Rai co-pilot
online · answers from live data
Highlights
The tools that pull their weight
A few of the screens that make Rai more than a POS — each one live in the demo.
See it, don't read it
The features, working — not just described
The parts that are hard to picture in a bullet list, shown as the product actually behaves.
Menu engineering
Every dish placed on a profit map
Popularity against real contribution margin — so you know which dishes are Stars to protect, Plowhorses to reprice, and Dogs to cut. A basic POS just lists what sold.
Same dish, two very different outcomes once the aggregator cut lands.
True per-channel margin
The dish that earns dine-in and loses on delivery
Once Zomato's ~30% cut and packaging come out, a profitable dish can quietly go negative. Rai shows the real margin on each channel, per dish — so you can price delivery differently or leave it dine-in.
Offline-first billing
Keeps printing when the internet drops
The most-cited complaint about cloud POS is that it stalls exactly at peak. Rai queues bills locally and syncs the moment you're back — nothing lost on a Saturday-night rush.
Live Floor · SLA timers
A timer on everything that can go wrong
Not just a table map — a countdown on a table seated too long, a ticket running late, a bill left pending. The floor runs on the screen, not on someone's memory.
Core, but not basic
Priced with the incumbents — but the floor works harder
The AI forecasting gets talked about, but it lives on the higher plans. Even the entry tier ships things a cheap POS doesn't — the operational wins that decide whether a busy Saturday runs smoothly.
Offline-first billing
Keeps printing when the internet drops
The most-cited complaint about cloud POS is that it stalls exactly at peak. Rai queues bills locally and syncs the moment you're back — nothing lost on a Saturday-night rush.
Basic POS: A cheap cloud POS freezes the counter until the connection returns.
Live Floor · SLA timers
A timer on everything that can go wrong
Not just a table map — a countdown on a table seated too long, a ticket running late, a bill left pending. The floor runs on the screen, not on a waiter's memory.
Basic POS: A basic POS shows table occupied or free, and nothing in between.
Pulled everywhere at once — no unfulfillable orders still landing.
One-tap 86
Kill a dish across every channel at once
Run out of paneer and one tap pulls the dish from dine-in, Zomato, and Swiggy together — no second tablet, no orders you can't fulfil still landing mid-rush.
Basic POS: Elsewhere you disable it thrice, on three different apps, from memory.
Day close · staff audit
Expected-vs-counted cash, and who changed what
Reconcile the drawer by payment method at close, with a full trail of every void, discount, and reopened table — per staff member. Day-end leakage stops hiding.
Basic POS: A cheap POS gives you a sales total and no way to see where cash walked.
Try them live in the demo — the offline drop and the 86 board are both interactive.
Why Rai
The part no incumbent POS does well
Billing and KOT are table stakes. This is what a kitchen actually gains from the AI layer sitting on top of them.
Demand forecasting that knows the Indian calendar
A generic trend line doesn't know an IPL final, a Ganesh Chaturthi week, or a monsoon Monday moves order volume 40–80% either way. Rai forecasts against an Indian event calendar built in, not bolted on.
Auto-generated reorder suggestions
When the forecast says Saturday needs 65kg of chicken and you're holding 38kg, Rai tells you today — with the reasoning attached, not just a number.
GST-aware costing, by default
Every item's slab, input-tax-credit exposure, and true margin — reconciled automatically instead of hand-patched at month-end by an accountant.
Built for peak-hour reliability
The most common complaint about incumbent POS is that it slows down exactly when a kitchen can least afford it — Saturday night, 12 tables deep. That's the metric this is designed around first.
On a ₹100 delivery order, after the aggregator's cut.
Every dish's true margin — even after Swiggy's cut
A dish can be profitable dine-in and quietly lose money on delivery once the ~30% commission and packaging come out. Rai shows the real per-channel margin on each dish.
A network, not just a restaurant tool
Connected kitchens pool demand for bulk rates none could get alone, see the network's real ingredient prices, and benchmark anonymously. Buying power a single-tenant POS structurally can't offer.
This week's auto-generated reorder suggestions
AI · reasonedChicken (boneless)
Order today38 kg on hand → suggest ordering 65 kg
Sat's IPL-night forecast alone needs 42kg — current stock runs out by Friday night.
Mozzarella
Order today9 kg on hand → suggest ordering 18 kg
Below reorder point; weekend pizza/garlic-bread orders trending up 12% vs last month.
Why this gets harder to copy over time
The screens are copyable. The data underneath them isn't.
Anyone can copy a forecasting tab. No one can shortcut yourkitchen's history — earned month over month, not coded once.
Forecast accuracy, one restaurant, since onboarding
64% → 93%
A new competitor starts every customer back at the left edge of this curve. The saffron gap is history they can't shortcut.
Public festival dates are free. The learned demand multiplier — per dish, per city, per cuisine — only exists after real transaction history.
Suppliers who see aggregate demand across many kitchens have to be won on both sides at once, not just sold a subscription.
Once ordering is wired through real suppliers, ripping it out is an operational decision, not a cancel-subscription click.
Honest comparison
Rai vs. the market leader
PetPooja does billing and aggregator sync well — that's why 60,000+ restaurants run on it. This is where the gap actually is.
| Capability | Rai | PetPooja |
|---|---|---|
| Billing, KOT & table management | ✓ | ✓ |
| Zomato / Swiggy order sync | ✓ | ✓ |
| Inventory & reorder point tracking | ✓ | ✓ |
| Offline-first billing that survives an internet drop | ✓ | ~ |
| Live Floor SLA timers (late-food / pending-bill) | ✓ | — |
| One-tap item 86 across dine-in + aggregators | ✓ | ~ |
| Day-close reconciliation + staff override audit | ✓ | ~ |
| AI demand forecasting, India-calendar aware | ✓ | — |
| Menu engineering (Stars / Plowhorses / Puzzles / Dogs) | ✓ | — |
| True per-dish margin after aggregator commission | ✓ | — |
| AI kitchen prep sheet from the forecast | ✓ | — |
| Automated reorder suggestions with reasoning | ✓ | — |
| GST-slab-aware item costing | ✓ | ~ |
| Designed against peak-hour slowdown | ✓ | ~ |
| Forecast accuracy that compounds per-restaurant | ✓ | — |
| Two-sided supplier demand network | ✓ | — |
| Collective purchasing (group buying across kitchens) | ✓ | — |
| Network ingredient price index | ✓ | — |
| Anonymous peer benchmarking | ✓ | — |
| Group P&L roll-up with outlier flagging | ✓ | ~ |
| Central menu/price push with per-outlet overrides | ✓ | ~ |
| Override approval workflows (discounts / voids) | ✓ | — |
| Automated UPI settlement reconciliation | ✓ | — |
| Vernacular AI co-pilot & morning briefings (Hinglish) | ✓ | — |
| Public changelog & status page | ✓ | — |
Based on publicly available feature listings and 2026 user reviews as of this writing — verify against current vendor docs before switching.
Concept pricing
Every plan unlocks its own slice of the live demo
This isn't just a price list — click a plan to open the dashboard with exactly that plan's features active, locked screens and all.
Founding memberLock today's price for 12 months
Join now and the rate you start on is the rate you keep for a year — or run a 30-day pilot first. The price rises for later cohorts, never for you.
Core
Run the floor
₹799/mo
- Billing, KOT & table management
- Live Floor — real-time dine-in & kitchen tickets
- Direct, WhatsApp & table-QR ordering — commission-free
- Virtual Brands — many delivery brands, one kitchen
- Inventory — stock, par levels & reorder points
- Zomato & Swiggy aggregator sync
- Covers & Peak Hours — covers-by-hour & table-turn
- GST e-Invoice — HSN, CGST/SGST, B2B-ready
- CRM & loyalty tiers
- Day Close — drawer reconciliation & audit trail
Growth
Most popularPredict the rush
₹2,999/mo
- Everything in Core
- Shift Commander — the AI agent fleet that proposes, you approve
- AI demand forecasting, India-calendar aware
- Forecast explainability & auto-reorder suggestions
- Prep Sheet — the forecast turned into mise-en-place
- Menu Engineering — margin × popularity, dish by dish
- Menu & 86 — live menu control + Margin Guardian by channel
- Reputation & Review Responder — replies drafted by sentiment
- Waste Watch — surplus-to-recipe suggestions
- GST-slab-aware costing & Hindi / vernacular KOT
Network
Compound the advantage
₹5,499/mo
- Everything in Growth
- Supplier Network + Procurement Intel — your prices vs the network median
- Aggregator Economics — true delivery margin + ad ROAS
- Delivery dynamic pricing, daypart-aware
- Win-back campaigns — approval-gated, over WhatsApp
- Benchmarks — anonymised peer comparison
- Staffing forecast + automated UPI reconciliation
- FSSAI compliance tracker
- Capital — data-driven working-capital offers
Enterprise
Run every outlet
Talk to us
- Everything in Network
- Multi-Outlet Console — cross-outlet view
- Central Kitchen — commissary production planning
- Chain Control + Anomaly Watch — variance & leakage flags
- AI-suggested inter-outlet stock transfers
- Roles, approvals & org-wide controls
- Multi-entity GST filing (GSTR-1 / 3B)
- Franchise royalty tracking + API & webhooks
- Priority peak-hour SLA & dedicated onboarding
Illustrative pricing for this concept — not a live offer. The demo plan switcher inside the dashboard lets you flip between all four at any time.
Manual now · auto at APIAggregator commission & ad data run on your uploaded payout/ads today; they sync automatically once the Zomato/Swiggy partner API is connected. Everything else is fully live.
Rai Saathi — for small dhabas, thelas & tea stalls
The segment PetPooja can't reach: a free, Hindi-first phone app with an udhaar khata, daily UPI + cash takings, and WhatsApp orders. No computer, no hardware, no subscription — the network pays, not the dhaba.
Bespoke
Shipped in 5 daysNeed a feature only yourrestaurant needs? We'll build it — live in five days.
Every kitchen does one thing differently. Tell us the capability no off-the-shelf POS has, and we scope, build and ship it into your account — custom, grounded in your data, yours alone.
- 1
Day 1Scope
A 15-minute call to pin down exactly what your kitchen needs.
- 2
Days 2–3Build
We build it into your account — grounded, tested, on your real data.
- 3
Day 4Review
You try it live and tell us what to tweak.
- 4
Day 5Ship
Live in your dashboard — a feature yours alone.
FAQ
Questions, answered straight
Is Rai a live product or a demo?
It's a working demo built on real, persisting slices — sign up and your data actually saves. The prices shown are illustrative for this concept, not a live billing offer yet.
Do I have to leave PetPooja to use Rai?
No. Rai matches the billing, KOT and Zomato/Swiggy table stakes, so you can evaluate it alongside what you run today. The difference is the AI layer on top — the agent fleet, forecasting and margin analysis.
Do the AI agents change anything on their own?
Never. Every agent only proposes. Nothing touches your menu, prices, supplier orders or customer messages without your explicit approval — each recommendation shows the exact source facts, and every approved action is logged in an audit trail.
Who owns my data?
You do. Your kitchen's own transaction history is precisely what makes the forecasts sharpen month over month — it stays yours, and it's the part a competitor can't shortcut.
Does Rai work for cloud kitchens?
Yes. Cloud-kitchen mode drops the dine-in screens and adds virtual-brand modelling — run many delivery brands out of one kitchen, on shared inventory and one KOT, with per-brand delivery economics.
Is it available in Hindi?
Yes. Hindi kitchen tickets are on the Growth plan (most POS handle regional-language print — this is table-stakes, not our edge). Where Rai actually differs is the vernacular AI layer: the co-pilot answers in Hinglish (“aaj ka sale?”) and your morning briefing reads back in plain Hindi — language in the intelligence, not just the print output. The free Saathi tier for dhabas is bilingual throughout.
Is payment secure?
Checkout runs on Razorpay, so card and UPI details never touch our servers, and every payment is verified server-side before anything unlocks. It's currently in test mode — no real money moves.
Feedback
Tell us what your kitchen needs
Rai is being shaped in the open. Every note here is read — say what's missing, what's confusing, or what would make you switch.