AI video moderation, with the evidence attached

Echosaw is multimodal AI video moderation software that checks each upload against a fixed harm policy — nudity, sexual content, violence, hate speech, and profanity — across video frames, the audio transcript, images, and documents. Echosaw is one engine: it ingests the media, analyzes it, scores it against the harm policy or rubric you enter, and summarizes what it found. It is built for teams that publish or review uploaded media and need a record of what was flagged and why.

Echosaw flags each issue, attaches timestamped evidence, and gates publishing: flagged media cannot be made public in the Echosaw Public Library. Your team reviews the evidence and decides what happens next, in the web app or in your own systems through the API. Echosaw analyzes media as it arrives — record in the browser, upload, or send files through the API — and returns findings as close to real time as recorded analysis gets. Need continuous feed ingestion? Contact us.

Most harmful video is harmful in one channel, not both. A clip can look ordinary and contain a slur in the dialogue, or be silent and show graphic violence. Echosaw treats video moderation as two checks on one file: the frames are checked for visual categories, and the transcript is checked for spoken ones. Both land in the same report.

Visual flags carry the time range where they occur. In the report, flagged moments show on the Timeline next to the scene they belong to, and content warnings appear in the Insights tab, so a reviewer jumps to the second that matters instead of watching the whole file. Click any finding and the clip plays from that moment.

When you try to make media public in the Echosaw Public Library, Echosaw checks its moderation flags first. Media flagged for profanity, nudity, sexual content, hate speech, or violence cannot be published, and the response lists the flags that blocked it. You can contact support to request an appeal.

AI content moderation services usually mean one of two things: a vendor that supplies human moderators, or software that scores content for you. Echosaw is multimodal AI moderation software: it gives your team the flags, the timestamps, and the evidence; your team makes the call.

If you need to check content against your own policy rather than a fixed harm list, load that policy into rubric scoring — the rubric is your policy, and the evidence is the media — and every finding comes back with a timestamp in the source. The same review model, applied to camera footage, is Echosaw for surveillance.