The working set in the first implementation, the capability behind it, and an honest account of what each addition costs to turn on.
Intacct is the golden source and always does the accounting. There is no option, no setting, and no client configuration that changes it.
Everything on this page adds operational capability. Not one item on it touches the books. Fuse records what happened on the floor and posts the consequence to Intacct; Intacct decides what that is worth and where it lands. Where Intacct holds a value, Fuse reads it. Where Intacct holds no value, Fuse refuses and reports rather than inventing one.
That is why the list below can be as long as it is without alarming a finance director. Adding maintenance scheduling or operator time capture to a manufacturer changes nothing about who owns the ledger.
Every movement, however it is entered, builds the same document and goes through the same posting code. There is no second implementation to keep in step.
Production against a bill of materials. Components off, finished goods on, at the cost Intacct holds.
Backflush decrease + run increase, one atomic post
Components out of stores into a work-in-progress warehouse, so consumption is visible before the job closes.
Inventory transfer
Warehouse to warehouse, carrying bin, lot and serial detail where the client uses them.
Inventory transfer
Balance, ledger, projected quantity and stock on order, off the same numbers Intacct holds.
Read-only reporting
Alongside these, a shop-floor screen carries the same three movement flows in a form a supervisor can use on a tablet at the machine: a scan, a quantity, one button. It builds an ordinary stock document and submits it, so a rejection from Intacct rolls the movement back and the operator reads Intacct's own words on the screen in front of them.
Leader Rubber's implementation deliberately took a narrow slice. That was their operating choice, not a limit of the platform. Below is what sits behind it, grouped by what turning it on actually costs — which is the number a consultant needs in order to scope a deal rather than promise one.
These are ERPNext capabilities that describe, schedule and control work. They move no stock and post nothing, so they are setup on an existing Fuse site — days, not a development project.
These change how work is modelled rather than how it reaches Intacct. The posting is written and proven; what changes is the shape of the movement flowing into it.
Each of these needs its matching Intacct posting written and proven against a live company before it ships. They are switched off today rather than half-built — a movement Intacct never sees is the one failure the product exists to prevent.
Nothing in tier three is promised as present. It is named so a consultant can price it, and so nobody discovers the boundary after a client has signed.
Fuse runs on ERPNext, so sales pipeline, customer service, projects, HR and payroll administration, document control and asset registers are all present on the same site, against the same customers and suppliers Intacct already holds.
Each one is a conversation about where the handover sits — and each one answers to the same rule. Operations happen in Fuse. The accounting consequence goes to Intacct, live, and Intacct decides it.
Item groups come from the product line tree. Decimal precision comes from the item record. Currency comes from the entity. Unit conversions come from Intacct's own UOM detail. Shipped defaults the company does not use are removed.
On a fresh install, nothing is typed by hand beyond the credentials.
Where something is genuinely ambiguous it is reported for a human rather than guessed. A wrong automatic guess is worse than no guess, and a consultant should never have to know which value Fuse chose on their behalf.
However much of the operation Fuse takes on, the answer to “who does the accounting” never changes.
That is what makes the list above safe to hand to a finance director.