Idomoo was founded in 2007 and is the first-to-market communication solution that combines the compelling power of video with personally relevant, individualized data. The Idomoo Video-as-a-Service (VaaS) platform integrates customer data into millions of automatically generated, yet seemingly tailor-made, professional video clips. Using these video clips, companies can change the way they interact and communicate with prospects and customers.
When Idomoo first began, the company decided to buy a few servers and host them with a third-party hosting provider. When customers began demanding better performance, scalability, and compliance with legal requirements, Idomoo quickly realized that buying hardware and managing servers was not the path for its growing business. Udy Ravid, Director of Product Management for Idomoo, describes what the company was looking for: “We were seeking a solution that could provide a significant performance improvement, as well as greater scalability and security. We wanted a cost efficient solution in an automated platform that could support all of the functions we needed to perform. We were also looking for a solution with advanced graphics processing unit (GPU) technology to grant us the ability to offer services on a larger scale.”
Why Amazon Web Services
“Being in the video mass production domain, our main requirements were significant performance improvement, scalability, and security. After testing a few alternatives, the GPU technology offered by Amazon Web Services (AWS) seemed to provide the best solution, fulfilling all requirements in an optimized and cost-efficient manner," says Ravid.
The Idomoo end-to-end solution, from the mass production of personalized videos to the video delivery is based on AWS:
Business logic and visual processing units: GPU-based instances for rendering personalized videos.
Database: CPU-based instance for saving data and tracking activities.
API/facade instance: CPU-based instance, receiving requests for personalized video generation via web services or other available interfaces.
Synchronization unit: Amazon Simple Queue Service (Amazon SQS) for coordinating and communicating between system units.
Storage: Amazon Simple Storage Service (Amazon S3) for storing and delivery of personalized video files.
The following diagram represents Idomoo’s process:
Figure 1. Idomoo Architecture on AWS
Ravid notes, “The programming language used for the solution is Java. In addition, other third-party video encoding tools are used. Being a fully automated system, several AWS SDKs are also used.”
Ravid describes the benefits of working with AWS: “As a business serving Tier 1 companies, such as major mobile operators and banks, we experience a high volume of video generation and hosting requests in short bursts. The elastic nature of AWS, combined with Idomoo’s elastic architecture, allows us to work efficiently without incurring local hardware and maintenance costs. AWS gives us the performance, scalability, and security at a low cost so we can exceed our customers expectations.”
By using Amazon S3, Idomoo can now store millions of video files inexpensively and quickly serve those videos to users. Another benefit of Amazon S3 is the ability for Idomoo to keep customer data within their respective continents, due to legal requirements.
Ravid notes that upgrading to AWS has benefitted the company in the following ways:
Savings of up to 30 times in video generation time. Due the time saved generating videos, Idomoo was immediately able to supply more videos.
Savings up to 8 times in video generation costs. Using the AWS cloud provided Idomoo better efficiency producing videos thereby reducing costs.
Savings in worker hours for system monitoring and support. With the reduction in video production time, Idomoo was also able to reduce the manpower needed to monitor and support internal systems.
Overall, says Ravid, “AWS offers the ultimate solution for a company like Idomoo, enabling it to provide scalable and secure solutions to Tier 1 companies. The GPU technology multiplies scalability, speed, and cost effectiveness.”