Veilio blurs explicit images and video on a reviewer's screen. They uncover what they need to see, when they need to see it.
How it works
How it recognises the material
Veilio uses AI models to recognise explicit images and video as they appear on screen. Another model writes the descriptions.
The models run on your own computers or servers. No image is sent to us or to any AI company, at any point.
Automatic detection can miss things. A box drawn by the reviewer stays blurred whatever the models decide, and if a model will not load or stops responding, Veilio blurs the whole window until it is working again.
If something goes wrong
If part of Veilio fails, more of the screen is blurred. A detector that will not load, a crash or an unplugged display all end with more covered than before. There is no fault that leaves a reviewer seeing more.
Where it runs
Records
Veilio logs each time something is uncovered: when, for how long and which part of the screen. It never logs what was on it.
The log can be collected by most security monitoring systems. Managers can see how much each person has been exposed to over a week or a month.
What it does not do
Veilio does not stop anyone seeing something they need to see, and it does not monitor productivity. The material still has to be looked at. Veilio reduces how often, and for how long.