How does moderation work in the Get Moments platform?

Get Moments can process UGC videos and photos from hundreds of thousands of fans in real time. Moderation had to be built into the core of our service, and it had to be fast, predictable, and configurable to the rules of each event, community, and sponsor.

But how exactly does it work, what are its limitations, and what can you as an event organiser expect when offering personalised recap videos for attendees via the platform?

This article provides an overview of moderation provided by the Get Moments platform. For an in-depth look at the tech behind it and details on what can be detected, download the full whitepaper below.

What we moderate: videos and feed content.

Moderation applies to all content uploaded to the Get Moments platform. When fans upload their photos and videos, that content is moderated and flagged or rejected outright if it violates policy. This includes uploads during the recap video creation process in the app as well as videos uploaded to the Feed either via the app or the web portal embedded on your website.

Our approach to moderation is based on three key pillars:

Pillar #1: Multi-layered defense 

AI filters run first, human reviewers use our built-in moderation tools to manage edge cases, and the community can flag anything that slips through.

Pillar #2: Organiser-led policy 

The first of two configurable avenues, event organisers and sponsors can decide what other content to ban, and even what to exclude from generated recaps.

Pillar #3: Per-event configuration 

Every activation and event is different. Filters can be added, removed, or retrained based on reference data provided by the organisers.

How moderation works: manual vs. AI-assisted 

Moderation can run in two modes, and each mode can be used on a per-event or per-activation basis based on your needs as the organiser.

Manual moderation

With manual moderation, every piece of content is reviewed by a person before it reaches either the recap or the Feed. This is made possible through two platform features:

1. Backstage review: your team can view content in Backstage, the platform’s admin console, and approve or reject one item at a time or by using bulk actions.

2. Community reports: also located in Backstage, this tab is where your team can review content flagged by app users.

AI-assisted moderation

In this mode, AI is the first filter that uploaded content passes through. Running in real time, AI will review every photo and video, while human input comes into play only to review whatever AI flags as uncertain, plus community reports in the same workflow described above.

The AI is from open-source recognition models pretrained on public datasets and fine-turned on real-event data collected from previous Get Moments deployments.

The filters can be turned off, replaced, or supplemented by custom filters based on data you provide us with ahead of time.

Since the filtering happens in real time, there’s no bottleneck to recap generation and the whole flow is faster and smoother for users.

For more on the AI moderation tech stack, what exactly can be detected and flagged, and our recommendation for the ideal moderation setup for your event type, download the whitepaper below.

Community reporting

AI is fast and scalable, but people are accurate and the community is vigilant. No one layer can work alone. So where does community reporting fit into our moderation setup? 

Anyone can report content through the app and include a reason in the report. That report then appears in Backstage in the Reported tab where the moderators can view and respond in real time. Moderators can approve, remove, or ban as needed.

Who does what: Get Moments vs. organiser responsibilities

The best moderation involves a combination of technology, policy, and people, and our setup is no different. Here’s what you can expect us to take care of, and what you should be prepared to provide for the most effective outcomes.

How violations are handled

When a user tries to use a clip that violates any applicable rule, policy, or law, the outcome is determined by you, the organiser. You can choose between two policies: permissive or restrictive.

Under a permissive policy, the clip can be uploaded to the user’s in-app media library, but will be excluded from any AI-generated outputs. The user will receive no message notifying them of the content violation.

Under a restrictive policy, the user will see a clear message alerting them that the clip cannot be added because it breaches law, community rules, or brand safety.

At the content and user level, there are several actions that can be taken.

Users can be warned, temporarily suspended, or perma-banned. Content can be banned and removed, silently excluded, or hidden from the Feed without being removed.

Want to know more about using Get Moments and its moderation features at your next event? Get in touch via hello@getmoments.com