Supply Chain Suite

Too much time preparing decisions,
too little making them

Predict. Plan. Perform. One decision layer across supply chain: Demand, Inventory, Manufacturing, Sourcing, Logistics and Fulfilment.

€1.4tnThe cost of decisions already taken
81%And it is taken in a spreadsheet
10%The decision still sits with a person
The problem

The bottleneck is no longer information. It is action.

The planner sees what systems miss. They know the customer who always orders late, the line that never hits standard, the part that goes short every August. None of it is in the master data. All of it belongs in the plan.

Buy one planning platform and you acquire one more spreadsheet per planner, because the spreadsheet is the last place a planner is still allowed to think. That is a verdict on the tools, not on the team. Nobody is asked to leave the place they think. Prediction, continuous rules, scenario tracking and a queryable graph sit underneath the work they already do, so the judgement stays and the assembly goes.

Why now

Supply chains have digitised the workflow, not the decision.

If the problem is decision quality, the platform must be organised around decisions, not workflows.

€1.4tn
The cost of decisions already taken

Held in excess working capital across the thousand largest European companies, fourteen percent of their combined revenue. Days inventory outstanding rose to 68.9, the highest in a decade, and the cash cycle has deteriorated three years running.

10%
The decision still sits with a person

Only one supply chain leader in ten will let AI make a critical decision without human review, and only thirty two percent are actively scaling it. Agentic AI sits at the very start of the analyst hype cycle, not the end of it.

81%
And it is taken in a spreadsheet

Four in five practitioners run sales and operations planning in Excel or Google Sheets. One in twenty uses the dedicated platform their company already bought. Half of the largest manufacturers still plan production on paper or in a sheet.

Sources
  • The Hackett Group, European Working Capital Survey, Nov 2025 (1,000 largest European headquartered non-financial companies)
  • RELEX Solutions, State of the Supply Chain, Jun 2026 (500+ supply chain leaders; the publisher is a planning vendor selling into the problem it describes)
  • McKinsey, supply chain survey, Nov 2023 (n=101; fieldwork more than two years old), 41% of advanced planning system owners still report excessive manual intervention
  • Gartner, Hype Cycle for Supply Chain Planning Technologies, Nov 2025
  • AbcSupplyChain, S&OP Survey, Nov 2025 (n=164 practitioners, 54 countries, 23 industries; self-selected audience of a training provider)
  • Qlector, How Leading Manufacturers Plan Production, Oct 2025 (largest Slovenian manufacturers, sample size not disclosed)
Six domain modules

Six domains on one decision layer

A supply chain that decides, not one that reports. Every module carries its own applications, built from the same four engines. Open a domain to see what runs inside it.

The advantage

Learn once, improve everywhere

Thirty problems across six modules, drawn from twelve classes. Eleven of the twelve are shared, so the second module costs less than the first. Every solved problem lowers the cost of solving the next one.

S1Parameter estimation and driftLead time, yield, safety stock, capacity and supplier reliability were typed in once and never tested against what happened.
S2Hierarchical and intermittent forecastingForecast at the level the decision is taken, when half the items sell four times a year.
S3Cold start and analogue transferForecast the thing with no history by borrowing from what resembles it and say how much of the answer is borrowed.
S4Multi-echelon propagationOne late component stops a line three sites away. The plan sees one level and stops there.
S5Constraint feasibility and sequencingCan the plan actually be run, on this capacity, in this order, with these changeovers.
S6Plan instabilityThe run moved four hundred orders, and nothing changed in the real world.
S7Exception triageRank what the run produced so a person only touches what costs money.
S8Plan to actual conformanceThe plan and the week diverged, and so did every human adjustment. Attribute the gap rather than argue it.
S9Event detection and effect estimationSomething happened or is about to. Price what it does to your plan, then learn from what it did.
S10Network traversal and exposureFollow the chain past tier one to the parts, orders and revenue at risk, with confidence on every link.
S11Allocation under scarcityFour hundred units, nine hundred units of orders. Decide by rule, not by who shouts loudest.
S12Scenario and attributionModel the intervention before you fund it, then prove the effect afterwards.
Changing the basis of competition

The plan was never the goal. The decision was.

Most platforms stop when the plan is produced. Prophesee continues until the decision is tested, actioned and proven. The best any category achieves today runs from storing the data to publishing the plan, and stops at administrative certainty.

We carry it further: predict the risk, surface the exception, evaluate the response, commit to a decision and prove the outcome. That is decision confidence, and it is the territory the incumbents leave empty.

Everyone optimises the plan. We optimise the decision.

Prophesee Supply Chain Suite
01Store the data
02Run the workflow
03Publish the plan
04Predict the risk
05Surface the exception
06Evaluate the response
07Commit to a decision
08Prove the outcome
Administrative certaintyDecision confidence
The platform

Four decision engines under every module

Foresight predicts what is about to go wrong, early enough to act on it. Pulse watches everything continuously and raises only the exceptions that matter. Horizon tests an intervention against scenarios before you commit to it. Nexus answers plain language questions across every system, then drafts the action.

The engagement

Eight weeks to a value you can see

Prove value before scaling. Start with one decision. Pick the module that hurts most, agree the metric and the baseline in week one, run on your data with your planners, and measure the result against the baseline we agreed.

Weeks 1–2

Design

  • Select the decisions that best validate the case
  • Capture today's baseline to measure against
  • Agree dates, owners and expected outcomes
Weeks 3–6

Develop

  • Connect the systems of record behind the decisions
  • Configure the suite applications on your data
  • Capture the rules that turn output into action
Weeks 7–8

Deliver

  • Validate results with end users against the baseline
  • Prove the agreed metrics with the evidence attached
  • Hand over the running system and plan the next modules
Customer outcomes

Proven, not promised

Procurement
How a FTSE 250 retailer saw vendor risk before it hit cash flow
Livespend vs budget

Real-time tracking of spend against budget with early warning for at-risk vendors and improved cash-flow management.

Choose one or more modules and start the evaluation

Pick the module where the pain is sharpest. Agree one baseline, signed off in week one. See the result, measured on your data, in your systems.