Pilot approval brief

# Pilot approval brief: Unnamed workflow

Organization: Not stated
Prepared by: Not stated
Generated: 2026-09-15T09:20:43.260Z

This brief separates requester assumptions from product measurements. Versalist makes no time-saved claim.

## 1. Problem

Not stated

## 2. Proposed pilot

Scope: One agent workflow, one challenge with fixed cases and rubric, and two or more agent versions compared with release reviews.

1. Set up: One challenge runs end to end for the chosen workflow and a baseline episode completes.
2. Compare: Each candidate version runs against the same challenge and gets a release review with a recorded ship, hold, or needs-work decision.
3. Decide: The manager exports the pilot metrics and an approval brief for the budget decision.

Roles and effort:
- Pilot owner (Engineering, AI, or product manager): About one hour per week. Chooses the workflow, records release decisions, and owns the approval brief.
- Integrator (One engineer on the agent team): One to two days during set-up, then under an hour per candidate. Connects the agent command, creates the challenge cases and rubric, and runs candidates.
- Technical reviewer (Security or platform engineer): One review session. Reads the execution and trace documentation and confirms data restrictions before any private data is used.

Commercial terms: Pilot duration, seats, run capacity, and price are not yet published. Request them through the demo form.

## 3. Expected value

Requester assumptions (not measured by the product):
- Releases per month: Not stated
- Review hours per release: Not stated
- Rework incidents per quarter: Not stated
- Hours per rework incident: Not stated
- Loaded hourly cost: Not stated
- Target hours from candidate completion to decision: Not stated

Derived exposure (assumption arithmetic, not savings):
- Review hours per month: Not stated
- Rework hours per quarter: Not stated
- Hours exposed per quarter: Not stated
- Cost exposed per quarter: Not stated

The pilot measures how much of this exposure the workflow actually changes. Do not present the exposure as a saving.

## 4. Cost

Pilot price or quote: Not yet quoted
Internal effort: Pilot owner about one hour per week; Integrator one to two days during set-up, then under an hour per candidate; Technical reviewer one review session.

## 5. Evidence

No release review attached. Attach one comparable review before requesting approval.

Pilot metrics

Not attached. Load metrics from the active organization once the pilot has reviews.

## 6. Success criteria

- Every candidate version in the pilot receives a comparable release review. Measurement: Review count equals candidate count in the pilot metrics. Source: product.
- Every release review ends in a recorded decision. Measurement: Decided reviews equal total reviews in the pilot metrics. Source: product.
- Score regressions are visible before a ship decision. Measurement: Reviews with a lower candidate percentage are listed in the pilot metrics with their decision. Source: product.
- Time from candidate completion to decision meets the requester target. Measurement: Median hours to first decision in the pilot metrics, compared with the target the requester states. Source: requester.
- Review effort per release stays at or below the requester baseline. Measurement: Requester records hours per review before and during the pilot. The product does not measure this. Source: requester.

## 7. Risks and open questions

- Hosting regions and encryption statements: Not yet verified (owner: Platform engineering and security).
- Subprocessor inventory: Not yet verified (owner: Platform engineering and security).
- Artifact retention and backup deletion: Not yet verified (owner: Privacy and business owner).
- Approved security contact: Not yet verified (owner: Security).
- Contractual service levels and certification evidence: Not yet verified (owner: Business owner).
- SAML or OIDC single sign-on: Not offered (owner: Platform engineering).

## 8. Recommendation

Undecided

Blockers noted at generation:
- Pilot commercial terms are unpublished.
- No release review evidence is attached.
- No pilot metrics are attached; value figures are assumptions only.