Demand prediction, adjusted for the Indian event calendar
AI recommendation · from your sales
Sat Oct 03: +40% orders forecast · 76% confidence
Above your typical day — averages 307 vs 220 overall Prep ~26 Hyderabadi Biryani and check stock ahead of the rush.
Forecast trained on your own sales
Weekday-seasonal model over 60 days of history (90 imported) · ~9 weeks
Sales Curator
Train the forecast on your history
Import a PetPooja sales export — the curator cleans it into a daily series so your forecast is real from day one, before Coal & Coin has weeks of its own orders.
Live signals feeding the forecast
The forecast isn't static — it adjusts to today's weather and the real event calendar.
Weather
🌧 Heavy rain expected, 6–9 PM
23°C · Dine-in typically −18%; delivery +12%
Event
IPL Qualifier · 7:30 PM
Biryani + wings +32%
Event
Ganesh Chaturthi eve (Wed)
Sweets + veg thali +80%
7-day order forecast
Saffron bars are the AI forecast; the faint bars are what a plain trend line would predict.
Mon, Sep 28
173predicted orders
-21% vs. trend-only baseline (220)
Usually quieter — averages 173 vs 220 overall
Tue, Sep 29
170predicted orders
-23% vs. trend-only baseline (220)
Usually quieter — averages 170 vs 220 overall
Wed, Sep 30
182predicted orders
-17% vs. trend-only baseline (220)
Usually quieter — averages 182 vs 220 overall
Thu, Oct 01
194predicted orders
-12% vs. trend-only baseline (220)
Fri, Oct 02
267predicted orders
+21% vs. trend-only baseline (220)
Gandhi Jayanti — a dry day in many states; alcohol-led venues typically softer (+8% vs your usual 247)
Sat, Oct 03
307predicted orders
+40% vs. trend-only baseline (220)
Above your typical day — averages 307 vs 220 overall
Sun, Oct 04
291predicted orders
+32% vs. trend-only baseline (220)
Above your typical day — averages 291 vs 220 overall
What's driving each forecast
Every number breaks down into named signals — nothing is a black-box output.
Jul 24
Base trend: 205
Jul 25
Base trend: 220
Jul 26
Base trend: 240
Jul 27
Base trend: 235
Jul 28
Base trend: 190
Jul 29
Base trend: 185
Jul 30
Base trend: 200
Tomorrow's prep plan — Tue, Sep 29
170 forecast orders (-23% vs a typical day) spread across your menu mix — prep quantities for tonight.
Hyderabadi Biryani
Main · normally ~34.3/day · 17% of your mix
26 plates
Butter Chicken
Tandoor · normally ~30/day · 15% of your mix
23 plates
Filter Coffee
Beverage · normally ~28.6/day · 14% of your mix
22 plates
Masala Dosa
South · normally ~25.7/day · 13% of your mix
20 plates
Veg Thali
Main · normally ~22.9/day · 12% of your mix
18 plates
Chicken 65
Chinese · normally ~18.6/day · 9% of your mix
14 plates
Paneer Tikka
Tandoor · normally ~13.6/day · 7% of your mix
10 plates
Gulab Jamun
Dessert · normally ~10/day · 5% of your mix
8 plates
Mozzarella Garlic Bread
Continental · normally ~8.6/day · 4% of your mix
7 plates
Prawn Biryani
Main · normally ~6.4/day · 3% of your mix
5 plates
This is on the Network plan
Staffing forecast and demand-shaping nudges unlock on Network and above.
See Network plan →Forecast accuracy, measured on your own history
Walk-forward backtest over 53 days of your sales — each day predicted using only the days before it, then scored against what actually happened. No hand-drawn curve; this is the model measured on you.
93%
overall accuracy · 91% in your first week
Buckets are chronological — earlier weeks had less of your history to learn from, so the curve is the flywheel, not a promise.
Auto-generated reorder suggestions
Forecast demand vs. current stock, with the reasoning attached — not just a low-stock alert.
Chicken (boneless)
38 kg on hand
Sat's IPL-night forecast alone needs 42kg — current stock runs out by Friday night.
Order 65 kg
Mozzarella
9 kg on hand
Below reorder point; weekend pizza/garlic-bread orders trending up 12% vs last month.
Order 18 kg
Paneer
14 kg on hand
Ganesh Chaturthi week typically shifts 15% of orders to veg thali — paneer usage rises with it.
Order 20 kg
Garam Masala Mix
3.2 kg on hand
2.4 days of cover against a forecast week with two demand spikes.
Order 6 kg
Suppliers see this demand signal too
Aggregate, anonymised demand across every connected kitchen — not just one restaurant's order.