Opportunity win probability
Predicted probability of winning a pipeline opportunity.
Part of Opportunities in the register catalog.
- Register ID
opportunity_win_probabilityUse this name to ask for the register. The API calls it a shelf ID.- Each record is
- An eventSomething that happened at a point in time.
- Status
- DeclaredDefined, and fills when a source produces its first record.
- Data
- OrganisationPrivate to each organisation's own workspace.
- Version
- v1Current version.
Connectors
Connected today
Demiton keeps this register current from these systems now.
No system fills this register yet.
On the roadmap
Systems in the planned connector rollout that can fill this register. Partial means the system holds only some of its fields.
None planned yet.
Fields
14 fields: 9 required, 5 optional.
Required
Every record carries these fields.
- p_winNumber
Model-issued probability that the referenced pursuit is won. The shelf's only measure, already declared
PREDICTIVE/float/probability.Example:
0.62- model_versionText
Version tag of the issuing model.
Example:
win_probability_v1- subject_opportunity_refText
Stable identifier of the scored pursuit.
Example:
opp:ITT242-caloundra-transport-corridor- training_as_ofDate and time
The transaction-time cutoff the model was trained to; the label set is only decided opportunities known at this instant.
Example:
2026-08-01T00:00:00Z- n_training_rowsWhole number
Decided (won/lost) opportunities the fit used;
MIN_TRAINING_ROWS = 5gates whether any prediction is emitted.Example:
24- n_winsWhole number
Positive-class examples in the fit.
Example:
9- n_lossesWhole number
Negative-class examples in the fit.
Example:
15- scored_atDate and time
When this prediction was produced.
Example:
2026-08-15T04:00:00Z- currencyText
Currency the scored pursuit is denominated in; today stamped
"AUD"unconditionally by the producer.Example:
AUD
Optional
Present when the source system records them.
- customerText
The customer the pursuit is for, where one exists.
Example:
Sunshine Coast Council- feature_valuesGroup of fields
The five bounded feature inputs to the score:
value_band(0–4),has_customer,work_type_known,sales_cycle_known,log_value.- valid_untilDate and time
Until when this prediction is the current one for its pursuit.
- brier_scoreNumber
Calibration error of the fit that issued these predictions, when a backtest was run.
Example:
0.17- accuracyNumber
Share of held-out decided opportunities classified correctly at the 0.5 threshold.
Example:
0.79
Example record
A sample from Bridgeworks Group, Demiton's demonstration company. The values are illustrative; the shape is exactly what the contract describes.
{ "accuracy": 0.79, "brier_score": 0.17, "currency": "AUD", "customer": "Sunshine Coast Council", "feature_values": { "has_customer": true, "log_value": 13.8, "sales_cycle_known": true, "value_band": 2, "work_type_known": true }, "model_version": "win_probability_v1", "n_losses": 15, "n_training_rows": 24, "n_wins": 9, "p_win": 0.62, "scored_at": "2026-08-15T04:00:00Z", "subject_opportunity_ref": "opp:ITT242-caloundra-transport-corridor", "training_as_of": "2026-08-01T00:00:00Z"}Use this contract
The contract is a standard JSON Schema (draft 2020-12) document, so any JSON Schema validator can check a record against it. Keywords starting with x-demiton- add Demiton's own meaning, such as units and requiredness, and validators ignore them.
https://docs.demiton.io/registers/opportunity_win_probability/v1.jsonShow the full JSON Schema
Loading the schema…
Talk to the team
Tell us what you run - your ERP, field, payroll and document systems - and we will show you the registers Demiton would keep current from them, on your data, in a 30-minute call.