Originally published October 1, 2026 · Updated October 10, 2026
Qwerty evaluation report template
Use this blank template to record your test setup, results and rollout decision.
Qwerty AI, Inc. · Local AI and agents for government, military and enterprise Mac fleets.1. Test scope and setup
Organization: ______
Group of test Macs and owners: ______
Workflow and source files: ______
Release and package hashes: ______
Model identity, source, hashes, quantization and settings: ______
Chip, RAM, storage, macOS, MDM and policy: ______
Network and retention rules: ______
2. Current process and test tasks
Current tools and settings used for comparison: ______
Representative inputs: ______
Quality scoring criteria and reviewer: ______
Number of runs, cold/warm conditions and concurrency: ______
Success criteria agreed before testing: ______
3. Results
Record measured quality, time to first useful output, task completion time, memory and storage use, behavior during longer runs and the range of results. Include failed runs as well as successful ones. Link each result to its test setup and supporting record.
4. Agent actions and offline use
Permitted source-file access: ______
Allowed and denied actions, and approval results: ______
Tasks completed with the network disconnected and data sent over the network: ______
Audit integrity, redaction, retention and export checks: ______
5. Installation and recovery
Clean installation and installation through device management: ______
Update on a small test group and approved rollback: ______
Recovery and approved diagnostics: ______
Support and handoff terms: ______
6. Decision
Expand, change the setup or stop: ______
Reviewer and date: ______
Blockers, known limits and people responsible for fixes: ______
Next qualification or retest date: ______
How to use the results
This blank template contains no customer performance measurements. Engineering check counts do not measure task quality or performance. Your results apply to the configuration and permitted uses you tested.