Cerul

Overview

Cerul is a powerful video search layer designed for AI agents, enabling them to "see" and understand video content. It allows for searching video based on meaning, encompassing speech, visuals, and on-screen text. This provides AI agents with the ability to process and act upon video information in a nuanced way.

Key Features

• Multimodal Search: Cerul indexes and searches across three key aspects of video content:
  • Speech: All spoken words are indexed and aligned with the video timeline.
  • Visuals: Searchable content includes slides, charts, demos, and whiteboard sketches.
  • On-screen Text: Text displayed within the video is also indexed for search.
• Single API Call: A unified API endpoint (/v1/search) allows agents to query across all these modalities simultaneously.
• Grounded Results: The API returns timestamped, relevant results that AI agents can directly act upon, including relevance scores and source links.
• Scalability: Built for scale, Cerul offers features like a Uptime SLA, low average latency, and fast search capabilities.

Use Cases

Cerul empowers AI agents by giving them the ability to understand and interact with video content. This can be applied in various scenarios, such as:

• Content Analysis: Automatically summarizing or extracting key information from video lectures, meetings, or presentations.
• Research: Finding specific moments or information within large video archives based on complex queries.
• Assisted Learning: Enabling AI tutors to reference specific visual or spoken content from educational videos.
• Media Monitoring: Tracking mentions or specific visual elements across video content.

Getting Started

Getting started with Cerul is straightforward:

1. Sign Up: Create an account to get an API key. A free tier is available.
2. Integrate: Use the provided API endpoint and documentation to integrate Cerul's search capabilities into your AI agent.
3. Build: Start building applications that leverage video understanding.

FAQ

What kind of content can Cerul search within videos?

Cerul can search across spoken words, visual elements like slides and charts, and on-screen text.

How does Cerul provide results to AI agents?

Cerul returns timestamped, grounded results with relevance scores and source links through a single API call.

Is there a free tier available?

Yes, Cerul offers a free tier to help users get started, with no credit card required.

What are the technical requirements for using Cerul?

You need an API key obtained after signing up and can then use the /v1/search endpoint with a simple curl command or through your preferred programming language.

Does Cerul offer documentation for developers?

Yes, comprehensive documentation, including an API reference and quickstart guides, is available.