AI Agents - Multimedia Autotagger Agent

Introduction

The Multimedia Autotagger Agent automatically tags and categorizes image, audio and video content for structured analysis across a wide range of use cases. Tags are user defined and can be fully customized, and the agent processes each row of media data automatically.

It can be used across a wide range of use cases, such as organizing media libraries, analyzing marketing and creative assets, monitoring brand and competitor content, or classifying social media advertising to identify themes, track trends, and quantify creative strategies.

The agent analyzes visual and auditory features without requiring manual annotation or keyword setup. It supports all major languages and emoticons, and includes advanced capabilities such as demographic identification, voice and image analysis, and flexible tagging formats.

Inputs to the Multimedia Tagging Agent

The Agent requires you to specify which column in your Redbird dataset contains the media assets to analyze. This column must include URLs that link directly to your image or video files.

These URLs can point to either:

Files hosted on your own cloud storage (for example, AWS S3 or Google Cloud Storage), provided the files are publicly accessible via a direct URL.
Assets collected using the Cloud Storage Collect app, which retrieves media from your cloud storage platform and creates a dataset with accessible links. Click here to view the guide on using the Collect app to build this dataset.

📘
The following file formats are supported by this agent

Images: png, jpg, jpeg, gif

Videos: mp4, mov, avi, webm

Adding Tags

Redbird allows users to create their own tags, which will appear as new columns in the output dataset. You will need to provide the following inputs to build a tag:

Name: This will appear as the column name in the output dataset.
Description: A description of what you want to tag for. This guides the AI on how to apply the tag.
Defined Values (Optional): A discrete comma separated list of values you want the AI to apply for this tag.
Examples (optional): Optional examples showing how the defined values should be applied. This is particularly useful for giving examples of specific text/labels/products that might appear in the creatives (for example, “Apple AirPods: Apple”, "Apple iPhone: Apple). You can provide these in any format, but a structured format such as the following is recommended: ["exampleValue1: tag1", "exampleValue2: tag2"]
Allow values note in list: When disabled the AI will only to tag content only using the defined values list. Enable if you want the AI to also generate tags that are not included in the defined values.

Agent Requirements

Below are the requirements (both mandatory and optional) for the agent to run successfully. These will either appear as input fields when using this agent through AI Agent Run.

Requirement	Description
Dataset to be Tagged	Source dataset for applying tags. Ensure it's added to the canvas and connected to AI DT, AI Agent Run or AI Chat node. For more info on collecting data into the platform see here.
Column to be Tagged	The column that contains the URL to the you wish to tag.
Context (Optional)	Describe the purpose of your tagging project and any relevant background that will help guide the AI. This should be a short paragraph in natural language. For example: “I’m a manufacturer of cosmetic products analyzing social media advertising from past campaigns. I would like to tag each asset with the relevant topics.”.
Tags	These are the tags you want the Agent to apply to your dataset. See the section above for more info on how to create tags.