1. Where AI may be used
- Classifying repair photographs or objects.
- Reading visible labels, model numbers, or part numbers.
- Suggesting likely part names, categories, or material options.
- Comparing request information with catalog records.
- Estimating whether more evidence or engineer review is needed.
- Assisting support, operations, or engineering triage.
2. AI can be wrong
AI output is probabilistic and may misidentify a part, miss a defect, infer a wrong dimension, or recommend an unsuitable material. Customers and Xahelo personnel should treat automated output as evidence to review, not as guaranteed fact.
3. High-risk decisions
Xahelo does not treat an AI result by itself as authorization to manufacture a safety-critical or restricted part, release production, pass quality inspection, or ship a regulated/high-risk item.
Human review or additional engineering evidence may be required when confidence, risk, geometry, use, or quality evidence warrants it.
If a privacy or automated-decision law becomes applicable to a Xahelo use of automated technology, Xahelo will add the required notices, access/appeal mechanisms, risk assessments, opt-out controls, or other safeguards before relying on that covered use. This notice does not state that every Xahelo AI-assisted workflow is a legally regulated significant decision.
4. Data used for AI processing
Relevant repair images, visible text, product details, and request context may be processed by an AI service provider when the feature is enabled. Do not upload unrelated sensitive personal information.
See the Privacy Policy for broader information about service providers, retention, privacy requests, and security.