Sanitised internal example

A release decision you can understand at a glance.

We ran the same eight synthetic tasks through two candidate routes. One completed every case. The other stalled three times. This page shows how that evidence supports a decision.

Method demonstration only, based on synthetic, non-personal data. It is not a customer testimonial.

Evidence ID: C-SAMPLE-2026-07System: internal AI workflowDecision: conditional proceed
Representative output · Not a certification
Decision

Use the selected route for controlled internal work. Hold the alternative.

The selected route completed the agreed test. The alternative must not be released until its timeouts and slow responses are resolved and the full suite passes.

What happened

8 / 8Selected route

All eight stored tasks completed.

5 / 8Alternative route

Only five tasks completed.

3Model timeouts

Three alternative-route runs never returned in time.

≈ 17×Median speed gap

3.30 seconds versus 55.81 seconds.

Quick read

Read this report in 60 seconds

  1. 1
    What we tested

    The same eight stored workflows, inputs and pass rules on two cloud routes.

  2. 2
    What we found

    A simple one-request check passed, but the full workload exposed three model timeouts and a large latency gap.

  3. 3
    What to do

    Keep the selected route for controlled internal use. Fix and retest the alternative before release.

Why this matters

One successful request looked reassuring. The complete workload told a different story. Construct makes that difference visible before users discover it.

Open the technical evidenceEverything below supports the plain-language decision above. It is included for engineers, reviewers and anyone who needs to reproduce the result.

01 · Test boundary

These results apply only to this version and environment.

System

Internal AI workflow prototype using one allowlisted language model through two candidate cloud routes.

Purpose

Create business drafts that always require human review. The system could not act on external services.

Candidate

Internal Vertex-only build tested on 22 July 2026 with server-controlled model and location settings.

Environment

Protected loopback prototype with eight stored workflows and synthetic inputs.

Retest after changing the model, route, prompts, dataset, production environment or ability to act externally.

02 · Full results

The complete comparison

MeasureSelected routeAlternative routeMeaning
Tasks attempted88Same workload
Useful and safe results8 / 85 / 8Alternative did not complete the suite
Model timeouts03Release blocker for the alternative
Median response time3.30 seconds55.81 secondsSelected route was about 17× faster
Longest selected response48.03 secondsSlow outliers still need production handling

03 · Actions

What the team should do next

  1. 1
    Fix alternative-route capacity and quota

    Repeat all eight cases only after the timeout cause is understood.

    High
  2. 2
    Improve one overly literal check

    The output expressed the right boundary in different words. The check was broadened and the change recorded.

    Medium
  3. 3
    Plan for slow outliers

    The selected route still needs queues, progress feedback, monitoring and incident handling before public production.

    Medium

04 · Traceability

What a reviewer can inspect

IDEvidence recordWhat it lets you verify
E-01Per-case evaluation fileInputs, duration, usefulness and safety result
E-02Route comparisonTimeout count and latency difference
E-03Versioned test runnerCases, checks and summary can be rerun
E-04Platform verification recordApplication and safety-boundary checks

05 · Limits

What this report does not prove

Not production approval

The test environment was an internal prototype, not the public production system.

Not exhaustive

Eight cases cannot reveal every possible failure or real-world outcome.

Not legal or security advice

This is not a penetration test, conformity assessment or regulatory approval.

Not transferable without retesting

The finding belongs to the tested version, cases, route and environment.

Plain-language glossary
Route
The technical path used to reach the model.
Timeout
The answer did not arrive within the agreed waiting time.
Median
The middle response time; half were faster and half were slower.
Test suite
The complete set of cases run together, not just one quick check.

Construct

Want this kind of answer for your AI agent?

Try the public demonstration first. To test a real system, we agree the decision, cases and safe access before any work begins.