Automate the first validation of eligible deliverables
As suppliers declare that elements have been delivered, Transparence verifies whether they are compliant with the contract's criteria. Then the contracting authority endorses the validation, or not.
Base validation on tangible evidence
Sometimes the contracting authority has no choice but to take the supplier's word for it, because it lacks the knowledge or the tools to make sure the deliverable meets expectations. Transparence changes that dynamic: the platform automatically verifies whether the deliverable matches the criteria written into the contract.
- Tie every payment to pre-established, validated proof of compliance
- Have deliverable compliance validated by your own validators
- Benefit from an efficient, ethical balance between technical proof and human accountability
An objective, fast verification endorsed by a human
The supplier submits the deliverable
It indicates where and how the deliverable is provided (URL, deployment, proof).
The platform launches the validation
The assigned validator or validators compare what was delivered to what the contract called for.
A compliance report is produced
Criterion by criterion, the platform indicates what is compliant, what is not and what is uncertain.
A human confirms
The machine has already validated compliance. The contracting authority reviews the report, then approves, requests a revision or rejects.
The payment is triggered
Once both validations are obtained, the payment can go out and everything is recorded in the registry.
Reconciling the machine's rigour with human judgment
The platform draws on the proven "two-person rule" principle, which requires that, for a critical decision, a single verification is not enough: a second, independent one is needed. In Transparence, the machine first validates compliance with the contract, then a human endorses that validation. This double confirmation, automated then human, secures every delivery before it is sealed.
The machine validates
The human confirms
Examples of validators
You build your own registry of validators, certified at the pace of your projects, in an extensible way. This lets you capitalize on existing tools rather than having to redevelop everything in-house.
Automated test analyzer
Automated test analyzer
Runs and validates the supplier's test suites: integration and end-to-end, for example.
Code coverage analyzer
Code coverage analyzer
Verifies that the test coverage percentage reaches the minimum threshold required by the contract.
AI visual validator
AI visual validator
Compares the deployed deliverable to the visual prototype approved at the start of the milestone, using computer vision.
HTTP performance checker
HTTP performance checker
Measures a service's response time and verifies that it meets the agreed performance threshold (SLA).
HTTP endpoint checker
HTTP endpoint checker
Verifies the availability and response of endpoints at the URLs agreed in the contract.
Kubernetes deployment checker
Kubernetes deployment checker
Confirms that the application is properly deployed in the target environment and that the service is active.
Cybersecurity validator
Cybersecurity validator
Relies on specialized platforms to confirm that a deliverable meets the planned security requirements.
Document certifier
Document certifier
Certifies a delivered document by computing its cryptographic fingerprint, guaranteeing it has not been altered.
Your custom validator
Your custom validator
A sensor on a physical structure, an inspection drone, a business validator: anything that can produce proof can be certified and plugged in.
You have questions? We have the answers
Here you will find the questions that come up most often about deliverable validation and the "two-person rule" in Transparence. If yours is not covered in this section, do not hesitate to contact us directly.
Concretely, when the assigned validators are launched, they can answer certain questions on their own: is the deployment active? Does the endpoint respond? Does the interface match the approved prototype? In short, the platform itself verifies that a deliverable meets the criteria written into the contract.
It is a proven security principle, common in very high-risk fields such as nuclear power, manufacturing and finance: a critical decision requires two independent validations.
In Transparence, these two verifications are complementary in nature: the machine validates the deliverable's compliance with the contract's criteria first, then a human confirms by endorsing that validation. The machine brings rigour and speed; the human brings contextual judgment. Depending on the contract, it is possible to ensure that no payment is triggered until both have confirmed.
That is the role of the visual agreement, designed for agile projects. During the project, the supplier submits an interface prototype that the contracting authority approves. That prototype becomes the reference against which the validator will automatically verify compliance.
Yes, because hierarchically, in terms of accountability and project management, human judgment must remain above the machine's judgment. If a contracting authority considers a non-compliance justified or minor, it can therefore force the approval despite the identified failure. The platform will then ask it to declare the risk level and a justification, which is then recorded as a validation override.
The goal is simply to make sure that an assumed exception is visible rather than buried, whether to train the machine or to learn from a mistake afterwards.
That is expected and accepted. In real life, there are limits to what can be automated. The platform clearly distinguishes what can be validated automatically from what cannot yet. For the rest, either a validator is added to the registry, or the element remains tracked through a manual human approval.
What matters is not that everything be automated on day one, but that every deliverable be explicitly tied to a verification method.
Validating faster without missing anything must become the norm
As digital projects grow in number and in the variety of elements to verify, the resources of Quebec's public sector remain limited. Automated validation removes part of the verification workload and leaves judgment to humans, where it truly counts.
For the contracting authority
Contract manager, project management office
- The ability to cover more projects, faster, with the same team
- Every approval rests on documented proof
- Tie every payment to confirmed compliance, not to an opaque decision
For the Auditor General
Auditor, independent oversight
- A compliance console that shows, deliverable by deliverable, what was validated and by which validator
- Major time savings on collection and validation
- The ability to go back in time and know exactly who approved what, and why
See the validation of a deliverable, from proof to payment
Rather than a generic demonstration, we can take a typical deliverable from your projects and show you concretely how the platform could validate it and what the human's role would be, in complete confidentiality.