How CareLens works

The browser checks that the upload is an image and prepares a smaller version before sending it for analysis. The server also validates the image format and size. This reduces unnecessary upload weight and gives the analysis pipeline a predictable input.

The current implementation sends the photo to an AI model through the Lovable AI gateway with a structured instruction set. The model is asked to describe visible cosmetic cues and produce practical skin, hair, body-care, diet and lifestyle suggestions. The output is structured JSON rather than a free-form page, and the application validates and normalizes the returned score fields.

A photograph cannot reliably measure nutrition, hydration status, sleep quality, hormone levels, vitamin deficiencies, disease, or other internal health conditions. Any score shown by CareLens is an app-generated heuristic, not a clinical measurement or diagnosis.

The response is organized into observations, routines, product categories, general lifestyle guidance, and a disclaimer. Missing sections are filled with safe empty defaults, and numeric scores are clamped to the application's 0–100 range. A hash-based server cache can reuse an identical analysis rather than repeatedly calling the AI service for the same image.

After a report is generated, the optional chat feature sends the report context and your conversation messages to the AI service to answer follow-up questions. The assistant is instructed not to diagnose conditions and to treat its output as general guidance.

An account is used when you want to save reports and track progress. Private application areas such as saved reports and account screens are separated from public editorial content and are not intended to be advertising pages. See the Privacy Policy for the current data-handling description.