PouchBook a demo
Industries · Business cashflow & forecasting apps

You sell cashflow foresight. We make it true.

Your product promises SMEs a forward view of their cash.

Flowcast · forward view

£4,715
#supplier-cycle · shifted +3 daysre-settled automatically

01

The problem

The feature you sell is the hardest one to build

Your product is cashflow visibility for SMEs, but most forecasts in the market still run on user-maintained inputs, spreadsheet imports or simple extrapolations.

Your forecast · user-maintained

last edited · 6 weeks ago
what it promiseddrift · rebuilt by hand
the feature you sellbreaks when a billing cycle changes

02

How it works

Your UI, your brand, our calculation

Step 01

Push categorised transactions

You send categorised business transactions to Pouch's endpoints as you receive them.

STRIPE PAYOUT · WK 27+£3,120.00

XERO · INV-882 PAID+£1,450.00

SUPPLIER DD · MERIDIAN−£640.00

POST /v1/transactions · 202

bank + accounting · if it's categorised, it qualifies

Step 02

The engine resolves the position

Pouch detects the frequency and classification of every recurring inflow and outflow, excludes internal transfers, and adapts automatically to changing billing cycles, supplier patterns and lumpy revenue.

A messy SME · read automatically

Revenue · payouts + invoicesweekly + ~monthly
Suppliers · shifting patternsdetected
VAT & insurancequarterly
wk 41 · seasonal dippredicted£820 low
#lumpy-cashflow · handled

seasonal, quarterly, shifting · forecast to the day

Step 03

One live number comes back

A single consolidated free-cashflow figure: income minus chosen bills, for today and any future day, across any forward period you define.

supplier DD shifted · monthly → fortnightly · detected

The forecast · re-settled

Live

User edits

0

Rebuilds

0

Support tickets

0

the maintenance your users churn over · gone

Step 04

You render it in your product

Your brand, your workflow, your customer relationship.

keep the product · buy the physics
◍FlowcastLive
New · this sprint

Forward cash position

£4,715

to the day · self-correcting · no user maintenance

Runway

7.2 months

Next tight week

wk 41 · £820

an API integration · not an engineering epic

03

The demo

What you will see in the demo

01A messy SME, forecast to the day
02Self-correction without user input
03Build versus buy, settled

04

FAQ

The questions your team will ask.

Forecasting is our core IP. Why would we buy the engine?
Your IP is the workflow, the insight and the customer relationship, and every part of that stays yours. The frequency and classification engine underneath is infrastructure: years of iteration you do not need to repeat. Keep the product, buy the physics.
Our data comes from accounting software, not bank feeds. Does that work?
If it is categorised, we can evaluate it. The pilot's sandbox stage establishes compatibility on one of your real customer datasets before anyone commits, so the answer comes from your data, not our assurances.
How accurate is the forward forecast?
The forward value is calculated from detected cycles, not guessed, and it resolves to the day rather than a coarse multi-month band. The pilot runs the engine on your own customers' history, so you judge precision on your book, not on our claim.
How heavy is the integration?
It is endpoints: push categorised transactions, get the numbers back. The Three-Stage Pilot runs sandbox integration, a live pilot, then reporting, and your lift is little more than a data export.
Next step

See it on your own data.

Book a demo, then judge the forward forecast on one of your real customer datasets.