Comment & DM moderation
Sence reads comments and DMs as they arrive across connected channels and acts against your rules. Harm is handled immediately while everything else stays open.
Moderation queue@grimjab_779:47 pmHidden
NegativePersonal HateViolent Language
@nineteenrows_back9:52 pm
Refs got that one right for once 😅
Decision
Hidden · Personal Hate · 9:47 pm
Classification and reasoning logged
Trusted by teams using Sence
How real-time moderation works
Thanks for having me London!
Connect your channels
Sence connects directly to Facebook and Instagram through the Facebook Graph API, and to YouTube through the YouTube Data API, so moderation happens where the comments actually live. There is no manual export and no batch processing.
Every comment read on arrival
Each comment and direct message is classified as it lands, against the taxonomy your team defined rather than a generic toxicity score. Classification keeps pace with arrival, so a spike does not become a backlog.
Hidden, flagged or left alone
Anything that breaches your rules is hidden before your audience sees it. Anything borderline is flagged for a person instead. Your team can hide, delete, notify or reply, and lock down comments on any post in one move when a pile-on starts.
What your team can do with it
Everything your team needs for moderation in one place.
@grimjab_77Hidden
NegativePersonal HateViolent Language
Hidden · Auto-moderated · 9:47 pm
Sence moderates in real time, hiding anything that breaches your rules as it arrives, on comments and on direct messages, 24/7.
Run fully automatic, semi-automatic or manual. Human review is available at any classification point, so your team takes the judgement calls and nothing else waits on them.
Spam and scams are the largest single category of harmful comment at 37% of everything Sence flags. In direct messages it is quarantined rather than left for someone to wade through.
When a pile-on starts on one post, your team can lock down comments on that post in a single move, without touching the rest of the channel.
Detection is not the end of it. Sence flags the risk and gives your team the actions: hide it, delete it, notify someone, or answer it.
Every decision is logged with its classification, the action taken and the reasoning, so nothing about a moderated post has to be reconstructed from memory.
Every comment and every message, on arrival
Comments on Facebook, Instagram, YouTube and TikTok, plus direct messages on Facebook, Instagram and TikTok. Facebook and Instagram connect through the Facebook Graph API and YouTube through the YouTube Data API, so there is no export step and no overnight batch.
Sence moderates comments and direct messages. That is the whole surface: the text your audience writes at you, on the channels above.
- Comments on Facebook, Instagram, YouTube and TikTok
- Direct messages on Facebook, Instagram and TikTok
- Automatic, semi-automatic and manual modes
- Every moderation decision is logged with its classification, the action taken and the reasoning, so your team can see what was hidden and why.
What it runs with
Protect
The whole job: what harmful comments cost you, and how Sence keeps a comment section open instead of switching it off.
Learn moreCustom Classification
What decides whether a comment gets hidden. Built from your own moderation decisions rather than an off-the-shelf toxicity score.
Learn moreTalent & Team Protection
The same moderation running on the personal channels of the players, staff and officials the abuse is aimed at.
Learn moreThe same classification pass that decides whether to hide a comment also reads what it says, which is why moderation andsentiment & themes come out of one product rather than two. If you want it in your own vocabulary, there are sector versions fornews and media and forsports.
Frequently asked questions
Comments on Facebook, Instagram, YouTube and TikTok, plus direct messages on Facebook, Instagram and TikTok. Every comment and message is read on arrival, anything that breaches your rules is hidden, and your team gets one place to see, moderate and respond across every channel.
Yes, on Facebook, Instagram and TikTok. Messages are read on arrival exactly as comments are, spam is quarantined rather than left sitting in the inbox, and anything that breaches your rules is handled the same way.
Comments are read on arrival rather than on a schedule, and abuse is hidden in seconds, even at the level a final or a breaking story pushes a comment section to. Nothing sits waiting for someone to open a tab at 9am.
Sence does not ship a default classifier, so no generic toxicity threshold is making that call. Yours is built from your own moderation decisions and your own policy taxonomy, which is what produces fewer false positives and fewer missed violations than a generic model. When your team does correct a decision, the correction feeds back into the classifier.
Yes. Every moderation decision is logged with its classification, the action taken and the reasoning, so your team can see what was hidden and why. That record is what you put in front of a legal team, a board or a platform when someone asks what happened on a post.