AI Processing Notice

Last updated: 2 October 2026 · Curiosbot Digital S.L., Zaragoza, Spain

CV Aligner uses AI models to understand Job Descriptions, find evidence in your CV and suggest improvements. This notice explains what is processed, by whom, and how we keep the output honest.

Which AI processes my data?

• Essential AI (Local AI): an open-source model (Qwen) running on Curiosbot infrastructure. CV content is processed on Curiosbot infrastructure and is not sent to a third-party AI provider.
• Smart AI: Google Gemini Flash via the Gemini API. Relevant CV and Job Description content is securely processed by Google.
• Expert AI: Anthropic Claude Opus via the Anthropic API. Relevant CV and Job Description content is securely processed by Anthropic.
• Analyze Match (requirement extraction and evidence mapping) uses the model shown on the analysis screen before you start. By default this is the local model.

We send only the structured content needed for each step (for example CV bullet points with their IDs), not your original file. Paid API providers process data under their API terms, which do not allow training on API inputs by default.

Automated decisions

CV Aligner does not make decisions about you. It produces suggestions and estimates for you to review. The alignment percentage is computed by deterministic rules from the requirement map, not invented by the AI.

How we prevent fabricated content

• Every suggested change references evidence from your original CV.
• A separate validation pass rejects new skills, tools, numbers, scope or seniority that your CV doesn't support. This happens even if the AI produced them.
• Where your CV only hints at something, we ask you to confirm instead of assuming.
• You review, accept, reject or edit every change. Your original CV is never overwritten.

Limitations

AI can make mistakes. Always read your CV before you send it. Scores are estimates and do not reproduce any employer's or ATS vendor's ranking.