R5Pilot · Striketing
AI Transparency Notice
How R5Pilot uses assisted extraction, what it cannot decide, and why human review remains required before imported data changes an alliance.
Last updated: 22 August 2026
What the system does
Screenshot workflows can extract candidate member names, ranks and operational metrics from uploaded images. The system normalizes values, matches candidates to the current alliance and highlights conflicts or low-confidence rows for review.
Data and processing boundaries
Processing is designed around a local worker strategy without a required external generative-AI API. Uploaded files are validated, bounded and treated as private tenant data. Temporary image payloads are removed after extraction or when the configured retention period expires.
Human oversight
An authorized alliance operator must review, edit, reject or approve candidate rows. Approved rows are still constrained by tenant authorization and domain validation. Missing or cropped screenshot data is not treated as proof that a member failed or should be removed.
Limitations and risks
Screenshots can be cropped, ambiguous, stale or incorrectly interpreted. Names may match the wrong roster member, numbers may be unreadable and game context may be missing. R5Pilot does not make disciplinary, employment, financial or other high-impact decisions for you.
Questions and feedback
Operators can inspect source batches and review decisions where the product exposes them, correct mistakes and request deletion or export through Striketing's official channels. Report unsafe, misleading or unexpected extraction behavior so it can be investigated.