Jump to…
AI Transparency Notice
Location: /ai-notice
Effective Date: April 27, 2026
Nodeline includes features powered by artificial intelligence. In the spirit of transparency — and to comply with Article 50 of Regulation (EU) 2024/1689 ("EU AI Act") and similar transparency rules — we explain here what our AI systems do, what they are not, and what their limitations are.
#1. What Our AI Does
| Feature | Function | Provider(s) |
|---|---|---|
| Résumé parsing | Extracts text from your uploaded PDF and structures it into fields (work history, education, skills, etc.) | Google Cloud Vertex AI (Gemini family) |
| Proposed positions | Suggests job categories from a predefined taxonomy that may match your experience | Vertex AI + rule-based scoring |
| AI Companion (dialogue) | Generates conversational text responses grounded in your résumé and prior dialogue | Vertex AI (Gemini); Anthropic Claude if enabled |
| Insights summary | Produces short text insights about your experience at your request | Vertex AI |
#2. What Our AI Is Not
- Not an employment decision system. The Service does not hire, screen, evaluate, rank, or decide anything about you for any employer. It is a self-use tool for individual reflection and planning.
- Not a licensed professional. The AI is not a lawyer, financial advisor, career counselor, recruiter, or medical professional, and its outputs are not advice from one.
- Not a source of ground truth. AI outputs can be wrong, incomplete, biased, or out of date. Verify independently.
- Not trained on you. We do not use your résumé, analysis, or private dialogue to train, fine-tune, or evaluate AI models. We require the same of our AI providers under contract.
#3. Your Inputs and What Happens to Them
- Your résumé text, profile metadata you optionally add, and dialogue messages are sent in real time to the AI provider's API.
- The AI provider processes them to return a response to us, which we return to you.
- We store inputs and outputs in your private account area so you can review history and continue past conversations.
- We do not store your inputs or outputs with the AI provider beyond the API call, except for short-term abuse-detection caches that the provider may maintain under its own policies. We select providers that offer zero-retention or enterprise no-training terms.
#4. Marking AI-Generated Content
Text generated by the AI Companion is labeled "AI generated" in the interface. If you copy AI text out of the Service and share it elsewhere (including in the Community), please keep the label and indicate it was AI-generated.
#5. Human Oversight
- You are in control of what you do with AI output. Nothing the AI suggests is a binding action on our part or anyone else's.
- A human member of the MathRhythm team reviews AI outputs periodically for quality and safety, and may adjust prompts, guardrails, or provider configuration.
- If an AI output seems harmful or clearly wrong, please flag it using the "Report" action within the conversation or email support@node.coach.
#6. Known Limitations
- Coverage: Our AI is tuned toward semiconductor career context. Responses about other industries may be less useful.
- Temporal drift: Underlying AI models have training cutoffs. Time-sensitive facts (laws, immigration rules, salary data) should be independently verified.
- Language: We support English and Korean primarily. Other languages may have reduced quality.
- Bias: Despite vendor efforts, AI systems reflect biases present in their training data. We encourage you to challenge outputs that feel off.
#7. Your Rights
You have the right to:
- Know when AI is producing content you see;
- Receive a copy of AI-generated outputs associated with your account;
- Delete dialogue history;
- Object to or request human review of any AI output that materially affected you (write to privacy@node.coach).
We do not use automated decision-making that produces legal or similarly significant effects about you (GDPR Article 22). The AI Companion is informational only.
#8. Updates and Governance
We review our AI features and this notice at least annually, and whenever we change providers, models, or core behaviors. Material changes are announced in the Service. The current list of AI providers and versions is maintained internally and available on request at support@node.coach.