What the AI analyses
- The photographs you select for a comparison or timeline insight.
- The journey type and the goal you set for it.
- The time between photographs.
- Photo quality hints such as brightness, blur and framing differences.
- Your notes and completed tasks, when you ask for next steps or a monthly summary.
What an observation is
An observation is a short, hedged description of visible difference — wording such as 'appears', 'seems' or 'looks'. It is a description of pixels, not of your health, and it never names a condition or recommends a treatment.
What confidence levels mean
Confidence describes the quality of the comparison, not the strength of your progress. Each observation also gives a short reason for its level.
- High: lighting, angle, distance and sharpness look consistent, so the comparison is reasonably reliable.
- Medium: minor differences in framing, lighting or quality may be influencing what is visible.
- Low: conditions differ noticeably — a changed angle, different lighting, blur or an obscured area — so the observation should be treated with caution.
Why results vary
Generative models are probabilistic. The same pair of photographs can produce differently worded observations, and small changes in lighting or posture can shift what the model notices. Consistent photographs, taken in the same place with the same framing, reduce that variation more than anything else.
Correcting or removing output
- Regenerate an insight to get a fresh reading of the same photographs.
- Edit or delete any note, goal or task the AI used as context, then regenerate.
- Delete an observation or a photograph from the entry at any time.
- Replace a low-confidence photograph with a better-lit one and compare again.
What the AI does not do
- It does not diagnose, screen for conditions or estimate risk.
- It does not judge the competence or quality of a clinician, salon, coach or treatment provider.
- It does not measure clinical outcomes or make predictions about your future results.
- It does not make automated decisions with legal or similarly significant effects about you.
Where processing happens
When you ask Journae for an AI observation, the selected photographs and the surrounding context (journey type, goal, dates, quality hints and, where relevant, your notes) are sent to a third-party AI provider through our AI gateway so the response can be generated. We do not use your photographs, notes or reports to train our own models. We cannot warrant how every upstream model provider handles data internally, so we only state what is within our control: we do not sell your content, we do not publish it, and we do not share it for advertising.
Reporting problematic output
If Journae produces output that feels unsafe, alarming, biased or wrong, email support@journaeapp.com with the journey type and the date of the entry. Please do not attach the photographs unless we ask. Reports like this directly shape our prompts and safeguards.