Cross-system mapping
Match identities, dates, units and authority across the systems involved in the decision.
Know what is likely to happen. Decide what to do about it.
Fable builds predictive Decision Engines for UK and US upper-mid-market and enterprise operators. Each engine combines production ML, cross-system evidence and financial attribution.
EvidenceERP, CRM and operating records joined at the cutoffForecastOutcome, timing and uncertainty estimated before actionDecisionNamed action, owner, deadline and expected effect
A forecast must change the decision.
Each engine handles one recurring operating choice. Models estimate uncertain outcomes. Rules exclude unsafe or impossible actions. Optimisation assigns limited time and capacity to the actions with the strongest expected result.
- 01Reconstruct the position
Join the records that describe the same project, product, customer or obligation at the decision cutoff.
- 02Forecast the outcome
Estimate what is likely to happen, when it may happen and how uncertain the forecast is.
- 03Compare available actions
Remove options that breach authority, policy, capacity or operating constraints.
- 04Choose the work
Rank the remaining actions by their expected operational and financial effect.
- 05Measure the result
Record what was approved, completed and achieved, then feed the outcome back into the engine.
Project Margin RescueReference run
Evidence spineStatusCutoff
ERP · project accountingAuthoritative08:42
PSA · time and deliveryReconciled08:38
Billing · WIP and ARCurrent08:40
Predicted outcomeMargin-at-completion distribution
DownsideForecast rangeUpside
01
Feasible interventionSubmit bounded EAC revision
Northbank Programme · owner PMO South · due 14:00
Expected effectProtect marginuncertainty shown in workbench
What each engine combines.
Production ML estimates uncertain outcomes. Cross-system mapping establishes the facts at the decision cutoff. Financial attribution records predicted value, realised value and finance recognition separately.
Production machine learning
Estimate the possible outcomes, their timing and the uncertainty around them.
Financial attribution
Keep predicted risk separate from action effect, then compare the result with the original expectation.
Decision Engines for material operating choices.
The catalogue covers commercial economics, revenue, cash, supply, workforce, operations, procurement, capital and risk. Two engines are shown in detail below. The full catalogue follows on the Decision Engines page.
Project Margin Rescue Engine
Determine which active project requires intervention now to protect margin, cash timing or delivery economics.
- Primary operator
- Finance, PMO and delivery leadership
- Cadence
- Weekly, with month-end escalation
Evidence5 systemsDecision grainProject × cutoffBoundaryApproval required
Harbour EastEAC driftSubmit forecast revisionDecision now
Lumen DeliveryBilling lagIssue eligible invoiceApproval
Alder ControlsCost exposureCorrect supplier itemMonitor
Range-to-Cash Trading Yield Engine
Choose the product, customer and branch actions that improve cash, contribution and service together.
- Primary operator
- Trading, category, finance and supply
- Cadence
- Weekly and event-driven
Evidence6 systemsDecision grainSKU × branchBoundaryApproval required
Alder Chilled · CentralExcess coverSuspend replenishmentDecision now
Juniper Pantry · NorthService exposureTransfer 48 unitsApproval
Elm Bakery · SouthWeak yieldHold through reviewMonitor
Trace every recommendation.
Each decision retains its source records, model version, constraints, alternatives, selected action and observed outcome.
Review the evidence01Production code executes the reference decision
02Models are compared with fixed baselines
03Every action has an owner and completion test
04Expected and realised value are recorded separately
Qualify the decision before committing to a pilot.
The Qualification of Opportunity Assessment checks the value at stake, the historical evidence and the operator's authority to act. A viable opportunity then moves to a measured pilot.
Paid QOA
Test the decision, available history, authority to act and value at stake.
Pilot
Run the engine against an agreed baseline using the client’s operating record.
Production
Deploy the approved models, workflow, integrations and fallback.
Stewardship
Monitor performance, retrain when evidence warrants it and account for results.
Operational software built around the business.
The Fable Systems Library covers CRM, scheduling, finance, accounting, quoting, billing, workflow, portals and integration. Bespoke Builds adapt that software to the client's operating model.
- Production software with defined users, records and controls
- Built around the client's data model, workflow and commercial edge
- Suitable as the workflow and evidence base for future Decision Engines
Fable Systems LibraryReusable product foundations
01CRM & portals
02Finance & billing
03Scheduling & workflow
04Integration platforms
Human approvalAuthority and permitted actions are explicit.
Tenant isolationClient records stay inside the approved boundary.
Deployment choiceCloud, private cloud or client-controlled infrastructure.
Bring us one recurring decision with a measurable financial consequence.
Tell us who makes it, how often, which systems hold the evidence and what a better decision would change.
Discuss the decision