Website video needs search functionality, obviously it does, otherwise we wouldn’t be able to visit our favourite sites and look for all the content that we haven’t been able to ingest through the new breed of TikTok-style (other socially-empowered content web platforms are also available) sites now proliferating across the web. But how do the sites we love to use know how to deliver the right content to us? They use image and video platforms with search capabilities and API connection capabilities.
Cloudinary is one such platform.
The Silicon Valley-based image and video company has now extended its core platform with a product it calls MediaFlows, which includes its work workflow generative Artificial Intelligence (gen-AI) engine known as FlowAI. Capable of natural human-typed language into executable workflow logic, a term we will explore and explain below.
How it works
With FlowAI developers can describe what they want a website’s video delivery functions to do (and what they want to build in terms of software code to deliver that functionality) in plain words – and FlowAI will automatically generate the required code.
Can we illustrate this someone abstract-sounding concept more clearly? Yes, we’re glad you asked. Imagine a user goes to their favourite news website and browses some news updates, reads a few articles and newsflashes and then also watches some video snippets. If (as we have all done whether for interest or by mistake) the user right-clicks on an item and selects ‘Inspect’, the source code that works to deliver that content is the stream of (incomprehensible to most) code lines that make up the visual media workflows that relates to how each media element needs to be treated in various scenarios.
That’s actually a lot of coding, right? MediaFlows automates the development of that code relating to the visual media elements.
Each workflow flow is composed of a library of Media Blocks. These are low-code widgets representing key media tasks. These widgets will connect to each other to program the behavior of a developer’s application, allowing developers (and other technical users) to design, deploy, test and scale custom media applications. With Cloudinary, all that happens in minutes (it used to be hours and perhaps days), while also integrating with Cloudinary APIs and services. Additionally, developers can save time by customizing flows shared by others.
Currently in beta, MediaFlows enables developers to easily manage, transform, and orchestrate visual media without writing code for developers responsible for building media-rich applications. The drag-and-drop interface and pre-built templates help automate a wide range of visual media use cases allowing developers to reduce costs and ship prototypes faster.
“Cloudinary [has demonstrated with this release] the practical adoption of generative AI – and MediaFlows is the latest Cloudinary product to harness the power of this technology to redefine what’s possible with visual media,” said Tal Lev-Ami, co-founder and CTO of Cloudinary. “Forget point-and-click. With FlowAI, developers and business technologists can transform ideas into production-ready automation with a few words. This means less coding – and more creating.”
Developer video functionality pack
The functionality scope here includes the ability to perform image and video transformations (such as background removal or recolouring) using Cloudinary APIs. The platform can connect and orchestrate workflows with platforms like Akeneo, Twilio, SendGrid and others.
Developers can use Cloudinary add-ons such as Amazon Rekognition (technology that works to deliver pre-trained and customisable computer vision (CV) capabilities to extract information and insights from images and videos) for image moderation and Google Video Intelligence tagging. It can also be used to build workflows for metadata management, tagging, cropping, and other asset management tasks.
The platform’s out-of-the-box templates, logic and integrations reduce coding time on visual media-related tasks by an average of 70%, accelerating development so teams can get to market quicker. MediaFlows is part of Cloudinary’s Emerging Business unit, a practice dedicated to incubating and accelerating the development of transformational new visual media technologies.