Privacy policy.

Last updated: 25 August 2026

Neteon Technologies operates this U.S. distributor website for YUAN High-Tech products. We help U.S. customers evaluate, purchase, and use YUAN capture, compute, and connectivity products. This policy explains what information we collect through this site and how we use it.

What we collect

When you contact us or request a quote, we collect the details you provide, such as your name, company, work email, phone number, region, product interests, and project requirements. Our web servers may also retain standard access logs used for security and site reliability.

How we use it

We use your information to respond to inquiries, prepare quotations, help select products, and support integration or OEM and ODM discussions. We do not sell personal information. We share it only with the YUAN manufacturer and service providers when needed to answer or fulfill your request.

Lead management

Contact and product inquiry submissions may be stored in HubSpot or another customer relationship management system. When email ticketing is enabled, the U.S. sales team and the customer may each receive a copy of the request for follow-up and recordkeeping.

Shortlists and cookies

Products added to a shortlist are stored locally in your browser and are not sent to us unless you submit an inquiry or call-list request. The current site does not set advertising or analytics cookies. If measurement tools are added, this policy and any required consent controls will be updated first. You can clear shortlist data from the shortlist panel or your browser settings.

ASK YUAN AI

Answer processing

When a question is within ASK YUAN's supported industrial-hardware scope, the question and recent in-browser conversation needed to answer it are sent to our AI inference service, OpenRouter, and a model provider. Greetings and clearly unrelated questions receive a fixed local introduction and are not sent for model inference. Model requests are restricted to provider endpoints that declare zero data retention for prompt and response content. The inference service may still process basic operational metadata such as timestamps, the selected model, and token counts under its terms. Neteon Technologies does not save full chat transcripts or IP addresses in the ASK YUAN learning database. Do not enter confidential, export-controlled, health, or other personal information in the chat.

Suggested questions

We count selections of curated prompts so the opening suggestions can reflect common needs. For repeated free-form questions, the review queue may hold a bounded normalized version, a one-way grouping fingerprint, a count, timestamps, and one short sensitivity-screened example, but only after ASK YUAN completes an answer. Questions containing common email, phone, web-address, credential, or instruction-abuse patterns are excluded.

Suggestions never publish automatically. An authorized administrator must approve an emerging question before it can appear publicly. Public popularity uses a rolling 30-day prompt count. Daily maintenance removes prompt-usage rows after 90 days, pending candidates after 180 days without activity, dismissed candidates after 30 days, and linked copies of approved candidates after 30 days.

Cached answers

The five opening FAQs use reviewed answers stored with the website and current public YUAN source paths, so selecting one does not require model inference. For future curated FAQs, we may retain one standalone generated answer, its public source paths, model label, content hash, and quality and coverage scores for authorized administrator review. Arbitrary free-form questions, follow-up conversations, low-coverage answers, and answers with invalid citations are not eligible. A reviewed cached answer is served only for the exact curated question and expires after 30 days unless it is reviewed again.

We may count anonymous answer-helpfulness feedback. To prevent false feedback, we retain a random response identifier, completion status, source count, and the privacy-safe retrieval observation described below for 90 days. A daily cleanup job enforces that limit. The record does not contain the question, answer, IP address, or visitor identity.

Coverage signals

Eligible questions may update a daily aggregate so we can find common information needs and gaps in search coverage. A server-keyed fingerprint identifies the aggregate. It contains a bounded intent category, canonical source paths, source and evidence-coverage counts, completion outcome, and helpful or unhelpful totals. It does not contain the question, answer, conversation, IP address, or visitor identity, and it is removed after 90 days.

These aggregates cannot teach or publish content automatically. An authorized administrator must create each persistent retrieval rule separately. A rule may contain only a reviewed label, bounded intent category, validated YUAN site paths, guidance, status, and audit information.

Abuse prevention

For abuse prevention, ASK YUAN may temporarily count requests using a keyed, one-way token derived from the request address. The raw address is not written to the AI database, and the token expires within minutes.

Retention and security

We keep personal information only as long as needed to respond to the request, maintain an appropriate sales or support record, or meet legal requirements. Access is limited to people and service providers who need it for those tasks.

Your choices

You may ask to review, correct, or delete the personal information associated with your inquiry. Submit a Privacy Request through the contact form and we will route it to the appropriate team.