Manufacturer of equipment for video content

Challenge

The Customer for the project is a company operating in the realm of video content creation and editing. They produce equipment, develop software, as well as hardware and software systems, and offer services related to video content creation and editing.

The Customer understood the necessity of innovative products with a wow factor to remain competitive in the video content creation and editing market. Therefore, they were always looking for ways to distinguish themselves from their competitors.

When the company began its collaboration with GDC Services, it already had a unique concept for a video studio that filmed from three angles, along with software that could edit these video streams.

The company offered this hardware and software system, along with a service package, to its corporate clients. They also provided services based on this system to the general public by installing these systems in shopping malls and other public places.

Thus, we needed to:

—  Conceptualize a unique new product for automatic video editing using AI in collaboration with GDC Services and experts of the customer's company. Capture this concept as a product development roadmap for the upcoming year.

—  Train and test AI models using video content filmed in the company's studios.

—  Develop a product based on the trained AI models, ensure its continuous development and quick delivery of value to the user — for automatic video generation from the filmed footage.

Solution

There were several mandatory conditions for the product's development:

1. The product should operate in a publicly accessible cloud, as well as within the software and hardware systems provided to corporate clients.

2. The product should transform the original videos into more dynamic, audience-engaging content, reduce the overall duration of the final video, but not omit key events and dialogues from the original content.

3. The product should be capable of handling various source material scenarios, such as single-camera presentations or multi-camera interviews.

To maximize the product's functionality, close collaboration was required between the customer's experts and the GDC Services team, who developed the product using Agile methodology.

The product is still being developed and improved. However, during its development phase, we have already implemented, tested, and released the key functionality — an event-time model built by Artificial Intelligence that identifies important events in the frame, people's emotions, unsuccessful takes, etc., and puts them on a timeline.

Next, AI, in conjunction with a layer of flexible business logic, determines the best way to edit and compile the best result into a single video track — deciding what to speed up, when to switch cameras, what to cut, and what to show in close-up.

Initially, the plan was to implement the necessary functionality in two stages. However, as these functions were being implemented, the combined team and the company owner shared new feature ideas and identified new categories of potential users for such a product. So far, we have completed three stages, and a fourth stage is planned to further enhance the product's functionality.

Thanks to the use of microservice architecture, the product is versatile in terms of infrastructure — it can be installed on a server within a corporate network for corporate clients, and it is also available as a public subscription service

Results

—  New product in the customer company's portfolio

—  Unique market offering that competitors lack.

—  Product sales have already generated revenues exceeding $150,000.