Definite is backed by Y Combinator

From control documentation to reviewer-ready workpaper.

Definite reconciles the population, executes each test attribute deterministically, routes incomplete evidence to reviewers, and delivers the workpaper with its evidence pack.

Bring one control

Connect, map, test, review, deliver.

1 · Connect

ServiceNow, Workday, Oracle EBS — read-only, with live read stats.

2 · Map

43 concepts, 65 field mappings, 3 source bindings — versioned.

3 · Test

Attributes A–F recognized from the test steps; population reconciled; seeded random sample of 10.

4 · Review

Verified failures separated from items with incomplete evidence.

5 · Deliver

Annotated evidence, evidence-pack download, 5-tab workpaper per retained run.

Missing evidence is not a pass. It is not a fail either.

ITGC-CM-01 · Change authorization · Q2 FY2026 · sample 10 of 214Engine conclusionCANNOT_VERIFY · 3 EXCEPTIONS + 1 REVIEW ITEM
ChangeAttributeFindingResult
CHG0040140A · AuthorizationApproval recorded after implementation startedFAIL
CHG0040234C · Segregation of dutiesImplementer approved their own changeFAIL
CHG0040436F · Emergency reviewRetrospective review completed lateFAIL
CHG0040391E · Test evidenceNo UAT evidence in the package — routed to reviewCANNOT_VERIFY
6 remainingA–FAll attributes held on retained evidencePASS

One item cannot be concluded: the evidence is incomplete, so the engine refuses to call the control tested. The missing package was requested from the implementer.

Lifecycle and verdict are separate columns. Untested controls sit honestly at “Not run.”

ControlStatusLatest result
ITGC-CM-01Change authorization & testingOPERATIONALFAIL · 4
ITGC-CM-02Emergency change managementOPERATIONALFAIL · 1
REC-01Account reconciliation prep & reviewOPERATIONALPASS
ITGC-LA-01Logical access provisioningIN DESIGNNot run
ITGC-OP-01Batch job monitoringSCHEDULEDNot run

Only executed controls carry a verdict. Every run is retained — one workpaper per run, filterable history.

Bring one control that is painful to test.

Control description, test steps, a population extract, and the evidence you have. We run it.

Show us the workflow