How Dermo's AI checks your skin

Most people who try a skin app worry about the same thing: is the AI actually any good, or is it just guessing? We built Dermo to be transparent about exactly how it works and where it can go wrong, because you deserve a clear answer when you are using technology to check something as important as your skin.

Three agents instead of one

When you upload a photo, Dermo does not ask a single AI for an answer. It asks three independent AI agents at the same time, each analysing the image on its own without knowing what the others think. Each agent produces a shortlist of possible conditions and a judgement on whether the spot looks benign or suspicious.

Once all three have given their initial answers, Dermo shows each agent the conclusions the others reached and asks them to reconsider. This second round works a bit like a medical consultation where three doctors compare notes. If one of them spotted something the others missed, the other two have a chance to update their thinking. The final answer is a consensus built from all three revised analyses, which makes the result more reliable than any single agent alone.

What the AI looks at

Dermo's AI has been trained to think the way a dermatologist does during a visual examination. That means paying attention to the morphology of the lesion (whether it is flat, raised, or nodular), the colour and whether it varies across the surface, the texture (smooth, scaly, crusted, or verrucous), the distribution pattern if there are multiple lesions, and the location on the body since some conditions favour specific areas.

For pigmented moles the AI has been trained to think through the classic ABCDE criteria — asymmetry, border, colour, diameter, and evolution — as one of several lenses it applies when forming its opinion. For non-pigmented lesions like warts, acne, eczema, or fungal infections, it relies more heavily on morphology and pattern recognition. In both cases the same three-agent consensus process applies.

Why you always see a confidence score

Every answer Dermo gives comes with a confidence level. If the AI is sure, you will see a high confidence score and a clear recommendation. If the AI is genuinely uncertain, the confidence drops and you will see a note explaining what is ambiguous.

We show you this number because it is dishonest to hide uncertainty. A medical tool that pretends to always know the answer is more dangerous than one that admits when it does not. You should weigh a low-confidence result differently from a high-confidence one, and the confidence score is your signal to do that.

What the AI cannot do

Dermo cannot replace a dermatologist, and we are not trying to. A smartphone photo gives the AI far less information than a clinician sitting in the same room as you has. A dermatologist can check the texture by touch, see the lesion under different lighting, ask questions in real time, and use a dermatoscope to examine deeper structures that are not visible in an ordinary photograph.

The AI also has blind spots. It performs best on typical lesions that resemble the ones in its training data. Rare conditions, unusual presentations, and lesions in locations that are hard to photograph (inside the mouth, under a nail, on the scalp) are harder for it to assess. Photo quality matters too, and a blurry, dim, or poorly framed image will get a less reliable answer.