Make videos with help from AI: GliaCloud uses GCP to create audio and video AI services, with one-click text to audio and video

The daily production of AI-produced videos is about 20,000, and the cumulative number of views is 1 billion.
The daily production of AI-produced videos is about 20,000, and the cumulative number of views is 1 billion.

GliaStudio is a tool that uses AI to help content creators mass produce short videos. It takes the film production process, which usually requires complex technology and is both time- and labor-intensive, and makes it simple and fast. That means that even people who don’t understand video editing technology can easily produce videos. Using AI technology, it applies natural language processing algorithms to analyze content themes and key points, automatically search for images, combine them into short videos, and turn articles into audio and video. A process that used to be done by outsourcing to professional audio-visual editors can be carried out by the enterprise’s internal staff after some brainstorming. This speeds up brand audio-visual production time, quickly responds to market topics and needs, assists content publishers, and increases the frequency of video updates.

Content management is quick, forceful, and precise: Attract cooperation with new media such as TikTok and Youku, with a daily production capacity of 20,000 videos

Because it can use news content, social media posts, real-time sports content, and statistics to create videos in a very short time, GliaCloud has cooperated with TikTok to produce thousands of Japanese and Korean videos every day. It has accumulated more than 100 million views and accelerated the overseas promotion of the TikTok app. During the 2018 World Cup competitions, it also cooperated with Youku to produce more than 5,000 edited videos of matches in real time through AI.


Flexible use of GCP Cloud Computing Services: GKE/GAE/GCE can easily deploy services and efficiently schedule resources

GliaCloud’s main business model is divided into two parts. One part consists of audio-visual advertising services for e-commerce, media, and advertisers. Because it has a large number of users, traffic is usually relatively high. The other part is the SaaS-mode audio-visual platform—GliaStudio, which has relatively stable usage.

As early as the beginning of 2018, GilaCloud was about to deploy services on the Google Kubernetes Engine (GKE) to accelerate the development of applications and ensure security. The most direct benefit is that its video production capacity has greatly increase—by 10 times, while the technical cost remain the same.

After that, in the face of high-traffic media audio-visual advertising services, GliaCloud also chose to run on the Google App Engine (GAE). Thanks to its high scalability and project control, even when there are huge surges in traffic for a major event, system performance remains unaffected.

Furthermore, for internal development, temporary projects, and test environments, going on Google Compute Engine (GCE) allows the technical team to shrink deployment time and introduce more new experiences in a short period of time. In this way, cloud resources can be flexibly adjusted and configured in response to changes in traffic and data storage demand. Application maintenance is easier with GCP.


Embrace new cloud technologies, deploy containerized applications faster and more securely, and optimize resource scheduling through Cloud Run

GliaCloud’s development team has always been open to experimenting with new technologies. For the sake of cost optimization, more stable system performance, and faster response time, it has recently begun preparing to gradually transfer services to Cloud Run. Cloud Run is a fully hosted service launched by Google that combines the flexibility and portability of containers, allowing users to run containerized applications on this serverless platform.
Cloud Run is responsible for the management, configuration, and expansion of all system infrastructure, which allows GliaCloud technical staff to spend their effort on writing code and developing functions without spending precious time and energy on the maintenance of the underlying system architecture. The integration of Cloud Run and GKE also allows GliaCloud to perform operations on the native GKE environment. That way, data migration, application deployment, and system settings are more convenient.

David Chen, founder and CEO of GliaCloud: “Our services encompass a wide range of technologies, from core AI algorithms to user-friendly UIs and stable AV advertising streaming services. Using the Google Cloud Native architecture allows us to focus on the R&D of product technologies instead of spending large amounts of effort on adjusting the system architecture. On top of that, iKala itself is also a heavy user of GCP. Using iKala to provide GCP technical consultation and service support, the fit between systems and human resources gives us a sense of security. “


Looking Ahead

GliaCloud is currently cooperating with more than a hundred high-quality media outlets in Taiwan to automatedly produce videos. In order to provide customers with a smoother operating experience and better audio-visual works, GliaCloud also looks forward to using more of GCP’s innovative technologies in the future to continue to strengthen product architecture and develop applications, use AI to help customers unleash their creativity, and promote the upgrading of digital content and the advertising industry.


GCE / GKE / GAE / Cloud Run


Company Introduction

GliaCloud was established in 2015. Its team is committed to improving the operating efficiency of the media and advertising industry’s audio-visual production through AI technology R&D. Its main product is GliaStudio, an automated AI audio-visual production platform, uses natural language processing, computer vision computing, and big data analysis technologies to quickly produce social media videos from news articles, sports events, statistical data, and other content in just minutes. Partners include such well-known international media platforms as Youtube, DailyMotion, TikTok, and well-known domestic and foreign advertising groups such as Publicis Groupe, Freakout and SalesFrontier. At present, the GliaCloud AI system has had more than one billion views of the video and audio it generated worldwide. The system supports five languages: Chinese, English, Japanese, Korean, and Vietnamese. It also continues to make breakthroughs and innovations in the types of supported video for production. It aims for being based in Taiwan while setting sights on the global market.

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Ut elit tellus, luctus nec ullamcorper mattis, pulvinar dapibus leo.

More Cases

Trend Micro

The golden ratio between cloud and on-premises: iKala reinforces the experience of using Trend Micro’s Google Cloud and turns to the DevOps revolution

Scroll to Top