Food & Beverage /Document Generation live field guide · 8 min

Batch Traceability & Recall Simulator

Maps lot and batch genealogy from raw-material receipt through production to customer shipments, tracing any lot forward to every affected product or backward to its source — with a simulation mode to run a mock recall and find traceability gaps before a real event.

Open the live demo

theater/demos/food-bev_batch-traceability.html · sandbox · read-only

Open

FIG. 1

The live demo, running on fabricated data. Open it to step through the full flow — every output is shown for a person to approve before anything happens.

How it would work

Reads a batch or ingredient-lot number against your receipt, production and dispatch records, builds the forward-and-back trace, and lays out every affected SKU, unit count and customer for a person to verify before any recall notice goes out.

Input 01

A lot number, and your records

One batch or ingredient-lot number — plus the receipt, production and dispatch records that link supplier lots to a production run, that run to finished SKUs, and those SKUs to the customers who received them.

Agent 02

Traces forward and back

Walks supplier lot → production batch → finished SKU → customer shipment in both directions, then totals the affected units, the customers impacted, and the distribution status — delivered, in transit, or still in the warehouse.

Output 03

A recall scope, for a person to verify

A draft recall scope — affected units, customer list, delivery dates and a per-customer notification — with every link shown, for a quality manager to check against the records and approve before any notice is sent.

Where it works well

It walks every link of the lot trace in seconds, in full, and shows its working.

Done by hand under pressure it is hours to days of cross-referencing — and the FSANZ Food Recall Protocol expects you to scope and notify fast.
Best for a quality or food-safety manager who runs mock recalls and real traces regularly — and for the annual traceability drill an audit asks for.
The recaptured hours go back into the judgement calls of a recall — risk assessment, regulator and customer communication — not the spreadsheet archaeology.

The slow, invisible cost in a recall is the tracing — joining a supplier lot to a production run, to the finished SKUs, to every customer shipment, across receipt, production and dispatch systems that don't talk to each other, while the clock runs.

Where it works badly

It is confidently wrong when the records are incomplete — and a clean-looking scope hides the gap.

Weak where lot links are broken — bulk decanted into unrecorded sub-lots, manual dispatch notes, or a supplier lot that maps to several runs it can't separate. It should flag the break, not paper over it.
If your genealogy lives in paper run sheets and dispatch dockets that nobody has digitised, the tool has nothing to walk and gives you less than your own records search already does.

The honest test

If you cannot say, right now, that every rework, repack and split since receipt was recorded against a lot — this tool makes your incomplete trace faster, not more complete.

Point it at a batch where a rework, a repack, or a split lot was never recorded, and it returns a tidy, professional recall scope that quietly misses product. A trace that looks complete and isn't is the worst outcome in a recall.

What it doesn't do — and shouldn't

It scopes and drafts. A person verifies and approves. That boundary is deliberate.

WHAT IT DOES

Surfaces the lot genealogy it walked, link by link

Shows the affected unit count, customers and distribution status

Flags every lot link it could not resolve cleanly

WHAT IT WON’T

Send a recall notification to a customer or retailer

Notify FSANZ, the state food regulator or a recall coordinator

Decide the recall class, or whether the food is unsafe

A recall carries obligations under Standard 3.2.2 of the Food Standards Code and the FSANZ Food Recall Protocol, and a missed unit is a public-safety and liability event. The decision to recall, the recall classification, and every notification stay with the accountable person — because the consequence lands on them and the business, not the tool.

What your data has to look like

Lot links recorded end to end — receipt, production, dispatch — as data you can query, not paper.

Typical readiness

across orgs we see, before the first job

Supplier lot capture at receipt

Needs shaping

Production genealogy — lot in, batch out

Needs shaping

Finished-goods to customer dispatch links

Usual weak point

Consistent lot and batch identifiers

Usual weak point

Distribution status per shipment

Needs shaping

The real first job

The genealogy is almost always the weak point — lot links live in paper run sheets, a dispatch spreadsheet, and an ERP that captures the PO but not the lot. Fixing how lot data is captured and joined end to end is usually the real first job — larger and more valuable than the tracing layer on top. Once the links are clean and consistent, every trace after that is fast and complete by default.

Right fit if…

You run mock recalls or traceability drills regularly, and want them to take minutes

Your receipt, production and dispatch systems already record lot and batch identifiers

You supply retailers who demand fast, documented one-up one-back traceability

You hold HACCP or SQF certification and the trace feeds your annual audit

Walk away if…

Your lot genealogy lives in paper run sheets and dispatch dockets nobody has digitised

Rework, repack and split lots are routine and rarely recorded against a lot

The same product carries different lot codes in receipt, production and dispatch

You need a tool that decides the recall class or notifies the regulator for you

Open questions

The worried-buyer questions, answered straight

It can return an incomplete scope — which is exactly why nothing is recalled or notified on its say-so. It walks the lot genealogy you’ve recorded, shows every link and unit count, and flags any lot link it could not resolve. A quality manager checks that against the records before approving the scope. The tool surfaces the trace; the person stands behind it — and the decision to recall, and the recall class under the FSANZ Food Recall Protocol, stay with them.

It works from structured lot links — supplier lot at receipt, production genealogy, and finished-goods-to-customer dispatch. Where genealogy lives on paper run sheets or in a dispatch spreadsheet, the trace is only as complete as what’s been digitised, and the tool can’t walk a link that was never recorded. Getting lot capture into clean, consistent, queryable form is usually the first piece of work — and the piece that pays off in every trace and drill after.

No. It removes the cross-referencing — the hours of joining lots across receipt, production and dispatch — so the manager spends their time on the judgement: is the food unsafe, what is the recall class, who must be told and how. The trace is still theirs to verify against the records and approve. The capacity it frees goes back into the parts of a recall only a person can do.

Current to the moment you trace. A recall scope is only as good as the dispatch and distribution status at the time you run it — what’s been delivered, what’s in transit, what’s still in the warehouse changes by the hour during an incident. Trace against yesterday’s dispatch data and you may notify a customer who’s already returned stock, or miss one who’s just received it. The honest test: do you know your dispatch status is live, right now?

Your supplier lots, production records and customer list are commercially sensitive, and the customer detail can be personal information under the Privacy Act. Any deployment runs against your own systems and data handling, not a shared pool — we scope where the data sits and who can see it as part of the build. The demo here runs entirely on fabricated data; Murray Valley Farms, Woolworths Epping and batch B-2026-0341 are not real.

No. It speeds the trace and drafts the scope, but it does not certify compliance. Your written recall system under Standard 3.2.2 of the Food Standards Code, the recall classification, and the notifications required by the FSANZ Food Recall Protocol stay with the accountable person. The tool gets the scope drafted; the person confirms it is complete and correct and runs the recall.

What it takes to build

4–6 weeks · 4 phases

Reused from template~65%

Bespoke to this skin~35%

stack · Claude · traceability/ERP connectors · graph view · review UI

What it would cost

Fixed scope, fixed price, fixed dates.

Bite-sized first piece

One contained change, low risk

Pilot build

Most builds land here

Embedded support

Scale on proof

Considering this for your operation?

The honest place to start is a bite-sized first piece — one contained change, low risk. Tell us where it hurts; we'll play it back, scope it, and show you what's possible.

Book a call How we work