OpsPanther forecasts parcel volume across your entire network — customers to sort hubs to delivery branches — and turns that forecast into staffing, truck and routing decisions far enough ahead to act on. Not a report you read after the week hurts.
Every sort centre keeps its own workbook. Nobody owns the network view. So capacity, labour and truck shortfalls surface in the week they happen — when the only options left are expensive ones.
Each hub forecasts itself, with its own assumptions and its own day-of-week habits. Nothing reconciles, so the network total is a guess nobody trusts.
A hub goes over capacity on a Tuesday in December. You find out Tuesday. Hiring, trailers and reroutes all needed lead time you no longer have.
Someone edits a hub's capacity or a customer's allocation, and every historical report shifts underneath you. There's no stable record to be held to.
This is the real model shape OpsPanther runs: a daily base, compounding growth, a day-of-week rhythm and a learned seasonal curve — checked against the capacity of the building that has to handle it. Move the sliders and watch the year respond.
Weekly volume against sortation capacity
Average parcels a day across the hub's customers today.
Compounded across the year, per customer.
How hard Q4 hits at its worst week.
What the building can sort in a week at current staffing and MHE.
Inside the peak week — daily shape
A weekly total hides the day that breaks you. Peak week averages fine and still puts Tuesday 40% over the door — which is why the engine models every day, not every week.
Freight arrives from other hubs too, and that volume is terminal — the receiving building delivers it and never sends it on. Plan labour against a hub's own volume and you will under-staff it every time. OpsPanther separates the two flows and sizes the building against the sum.
Where the volume comes from, and what it means for the floor
Originates here, from this hub's own customers.
Trunked in for this hub to deliver. Terminal — never re-sent.
Own volume that belongs to another hub's delivery area.
terminal—Once volume is forecast per lane, trailer fill becomes arithmetic instead of instinct. OpsPanther prices every lane, shows which runs can be consolidated, and flags flows big enough to ship direct and skip a sort entirely.
What better fill is worth on a single lane
What your trailers actually leave with today.
| Lane economics | Today | At target fill |
|---|---|---|
| Loads per week | — | — |
| Average trailer fill | — | — |
| Weekly carrier cost | — | — |
| Cost per parcel | — | — |
| Empty seats bought | — | — |
Every screen answers a decision somebody has to make this week: how many people, how many trailers, which lane, which building.
Allocation changes over time, compounding growth, day-of-week rhythm, a learned seasonal curve and holiday effects — customer, hub and branch in one model.
Separates a hub's own volume from freight received from other hubs, so labour planning matches what each building actually handles.
Hubs and branches heading over capacity are flagged well ahead, and peak season is sized a quarter out — while hiring and trailers are still available.
Lane utilisation, consolidation opportunities, and flows big enough to ship direct and skip a sort — each priced with real carrier cost per load.
Forecast measured against imported actuals, with calibration that learns each customer's rhythm instead of applying one network-wide fudge.
Completed weeks are frozen into immutable snapshots. Change anything today and last month's reported numbers stay exactly as reported.
Hubs, branches, customers, allocations and transit. Once — then it's a living model, not a rebuild each quarter.
Real volume flows in and the engine calibrates per customer: growth, day-of-week shape, seasonality.
Staffing, trailers and lanes sized from forecast daily volume, with shortfalls flagged before they bite.
A completed week is photographed and locked. History stops moving; the future stays editable.
Third-party logistics operators and parcel carriers running a multi-building network — several sort hubs, delivery branches beneath them, and customers whose volume is split across sites. If one person is holding the network forecast together in a spreadsheet, that's the job this replaces.
A dashboard reports what happened. OpsPanther projects what will happen and sizes it against the capacity you actually have — then holds itself accountable by measuring the forecast against imported actuals. The output is a decision, not a chart.
Your network structure and a history of actual volume. The more history, the faster calibration learns each customer's growth and weekly rhythm. You can stand up the structure and forecast from a daily base on day one, then let accuracy tighten as actuals arrive.
No — and that's deliberate. Completed weeks are frozen into immutable snapshots, and historical views read the snapshot rather than recomputing. Edit a hub's capacity or a customer's allocation now and past reports stay exactly as they were reported. Freezing and unfreezing is admin-only and fully audited.
Yes. Multi-tenancy is enforced in the database itself with row-level security, not in application code, so every query is scoped to your account at the lowest level. Roles and invitations are managed by your own admins.
They use the real model shape — the same base × growth × day-of-week × seasonal structure, the same two-flow sortation identity, and the same fill-and-cost arithmetic the product runs. They're driven by your slider inputs rather than a live account, so nothing here touches real data.
OpsPanther is live in production today. Open the app, or talk to us about your network and we'll walk through it with your own volumes.