Our story
Built by someone who has done serious moderation
modwall wasn't born in a conference room. It came from on-call shifts, midnight reports, and burned-out volunteer moderators — in a community of ~300 000 users.
What moderation looks like from the inside
When you run a community of hundreds of thousands of users, moderation stops being a "feature" — it becomes a daily reality. Someone has to review every report. Someone has to decide whether a photo breaks the rules. Someone has to be on call when, at 11:47 p.m., something lands on the forum that shouldn't survive a single minute.
That "someone" is usually volunteer moderators — the most underappreciated group on the internet. They review hundreds of items a day, including ones no one should ever have to see. They burn out. They leave. And the liability — legal and reputational — stays with the platform operator regardless.
What scale taught us
Three lessons that directly shaped modwall:
Machines for the routine
90% of cases are obvious — those should be caught automatically, right at upload. People are too valuable to review everything.
Humans for the edge cases
Borderline cases require context and conscience. That's why the human-in-the-loop queue has been in modwall from day one, not as an add-on.
A defensible trail
When a dispute arises — with a user, a lawyer, a regulator — whoever has the record wins: what, when, and why. Hence the immutable audit log.
That's why we don't sell an “AI model.” We sell what we needed ourselves for years and couldn't buy at a reasonable price: reliability, compliance, and peace of mind — at a price a small platform can afford.
Become one of the first documented deployments
modwall is a young product built on old scars. Our first customers get something no big platform can offer: direct contact with people who moderated themselves, a say in the roadmap — and a spot on this page once your deployment becomes a success story we will tell together.
Who we work for
Anywhere users upload content and the operator is responsible for the consequences:
Marketplace
Filter listing photos and seller avatars before they reach the catalogue — without building your own moderation pipeline.
Classifieds sites
Thousands of photos a day assessed automatically, with a manual review queue only where it is really needed.
Forums and communities
Scan attachments and avatars at upload time — moderators get only the borderline cases, not the whole inbox.
Dating
Profile photos checked before publication — protecting your users and your app's reputation from the first upload.
Comments and UGC
Images and text in comments and reviews under control — without building your own moderation pipeline.
Apps with avatars
Anywhere a user uploads a profile picture — integrate with the widget in minutes, with no backend changes.