Built for seasonal kitchens

The prep list

that writes itself.

Saltline learns your restaurant's patterns — weather, tides, local events, your POS — and generates a precise morning prep list before your crew walks in. Less waste. Fewer 86s. No spreadsheets.

Morning prep list

Saturday, July 19

ML · 6:04 AM
Rain forecastYork DaysNo school

New England clam chowder

Rain + weekend

10 qt

↑

Lobster bisque

Peak season

8 qt

↑

Lobster rolls

On track

52

→

Clam strips (breaded)

Slow Tuesday

18 lb

↓

Based on 14-day history + signals

Toast connected

Forecast accuracy

±9%

Accuracy — last 14 days

Waste saved this week

$110

↓ 18% vs prior week

Rain forecast — +23% chowder volumeYork Days weekend — +18% total volumeHigh tide at 6 PM — dinner push runs earlySchool vacation week — +30% lunch coversLobster Festival — dinner 86'd three nights straightNor'easter incoming — pull tonight's raw barJuly 4th weekend — 3× normal lobster rollsPost-Labor Day — coast clears, locals returnRain forecast — +23% chowder volumeYork Days weekend — +18% total volumeHigh tide at 6 PM — dinner push runs earlySchool vacation week — +30% lunch coversLobster Festival — dinner 86'd three nights straightNor'easter incoming — pull tonight's raw barJuly 4th weekend — 3× normal lobster rollsPost-Labor Day — coast clears, locals return
0%
Reduction in food waste
Average across restaurants in their first full operating season
$0
Monthly savings per location
From reduced over-prep, spoilage, and last-minute ordering
0 min
Saved on morning prep daily
Time recovered from manual guesswork and prep-list disputes
0%
Forecast accuracy
After one complete operating season of data collection

Platform

Everything you need to run tighter.

From the moment you connect Toast, Saltline starts learning. By the end of your first season, it knows your kitchen better than any spreadsheet ever could.

ML prep forecasts

Your restaurant's own model — trained on your sales history, local weather, tides, and events. It updates every night and sharpens with every season.

Morning prep lists

Generated before service starts, updated intraday with same-day pace data. No printing, no whiteboard, no disputes about whose number is right.

Barcode scan & count

Scan any product UPC to build your inventory catalog. Morning counts take 6 minutes, not 25. Works offline, syncs when you're back on WiFi.

Inventory & waste tracking

Barcode scanning, yield inference, and quantity learning from your actual depletion data. The system backs out how much you really use per cover — automatically.

Alerts & kitchen comms

Instant alerts when forecasts shift mid-day. Prep status visibility across the kitchen without radio chatter or group texts.

Intra-day nowcast

Live POS data from Toast feeds the model throughout service. If lunch is running 30% over projection, the prep list updates before dinner starts.

How it works

From connection to confidence in one season.

Connect

Link your Toast account. We pull your full order history — items, modifiers, quantities, timestamps — and start building your model immediately.

Learn

The ML pipeline trains on your sales against weather, tides, local events, and day-of-week patterns. Cold-start predictions begin on day one.

Prep

Every morning, your team gets a precise list. Every night, the model updates. Every week, it gets sharper. After a full season, accuracy typically exceeds 94%.

Track

Count inventory with your phone. Log waste with a tap. The system learns your yields from your actual depletion — not from numbers you type in.

Why coastal New England

Built by someone who's worked the line in York Beach.

Seasonal restaurants on the New England coast face pressures that generic software was never designed for. You do three months of revenue in the summer. A cold snap in July, a road closure on 1A, a canceled fireworks show — any of these can tank a Saturday that was supposed to carry the week.

I've spent seven years in kitchens on this coast. Saltline is the tool I wished existed.

"Know your line." — how much you can make, how much you need, where the waste is going.

Seasonal revenue swings

Your model is trained on your seasonal rhythm, not a national average. Week 1 of tourist season is weighted differently than week 1 of October.

Event & weather sensitivity

York Days, Hampton Beach concerts, Ogunquit Fest, school vacations, nor'easters — Saltline tracks the signals that move volume on the coast.

Small team, thin margins

No enterprise license. No dedicated IT. Saltline is built to run on a phone or tablet with zero training beyond scanning a barcode.

Off-season continuity

The model doesn't go dark in November. Off-season data is what teaches it why December looks nothing like August.

Your next season, smarter.

It takes about 15 minutes to connect your Toast account. By your first service, you'll have a prep list.

Get started freeBook a demo