线条科技

Baidao Helps Line Technology with Overseas Marketing

Customer background

Line International is a mobile internet company integrating mobile advertising placement and investment & financing. It has been deeply engaged in the overseas marketing industry for 7 years. With high – quality global media resources, the application ability of large – scale advertising data, and a team of senior experts, it provides effective marketing solutions for over 10,000 enterprises going global.

On the one hand, it provides a one-stop advertising platform and solutions for global advertisers. On the other hand, it offers efficient monetization services for a large number of mobile developers. It has strong media resources across all channels, a professional overseas service team, and is driven by technological innovation. The industry’s first intelligent investment SaaS helps advertisers reduce costs and increase efficiency. It provides customers with efficient, simple, and results-driven Super AI marketing tools, meeting the content marketing and creation needs of overseas enterprises in various aspects.

Business requirements and challenges

With the launch of the Super AI product, enterprises can significantly shorten the creation cycle, remarkably reduce labor costs and production expenses through automated content generation, enabling marketing teams to focus more on strategy formulation and effect optimization. However, with the increase in the number of users, the demand for computing resources has grown. Super AI needs to have the ability to dynamically adjust computing power resources according to real-time loads to ensure stable performance even during peak periods and avoid resource waste during trough periods.

With the accumulation of user data, the Super AI model needs to be continuously optimized and iterated to improve the quality and personalization of content generation. This includes regularly updating the training dataset and adjusting the model parameters. The training and deployment of the model require a large number of GPUs, and the dataset used for training the model needs massive data support, which means that a data storage and management platform with sufficient capacity is required.

The security and compliance of user data are of utmost importance. With the increase in the number of users globally, the platform needs to meet the data compliance requirements in various regions and ensure compliance with relevant data privacy regulations.

Solution

The architecture design diagram is as follows:

Powerful computing resources: Google Cloud offers various high-performance computing resources, such as GPUs, which can meet the training and inference requirements of Super AI models. It also supports on-demand expansion to reduce costs.

Massive data storage and management: Google Cloud offers various storage solutions, such as Cloud Storage and Cloud Filestore, which can securely and reliably store and manage the data of Super AI, and provide a reliable data storage platform.

Global infrastructure: Google Cloud has an infrastructure that spans the globe, which can help Super AI achieve low latency, high availability, and global reach, meeting the needs of users in different regions.

Security and Compliance: Google Cloud provides robust security and compliance safeguards to protect the data and system security of Super AI and comply with various industry standards and regulations.

Use the product

• Compute Engine
• GPU
• Filestore
• Cloud Storage 

Customer benefits

Note: [All the following statistical data are from the statistical results of the customer background]

• By using Google Compute Engine (GCE) and combining strategies such as preemptible virtual machines and sustained use discounts, customers have reduced the cost of infrastructure by approximately 40% compared to building their own data centers previously.
• Improve the network connection stability by % through Google’s high-quality dedicated network.
• By using Google Cloud’s file storage service (Filestore) and object storage service (Cloud Storage), and selecting the appropriate storage category based on the access frequency of different data types, the storage cost has decreased by 25%.
• With the help of Google Cloud’s high-speed network infrastructure spread across the globe, the average latency of the Super AI platform has been reduced by 30%, and higher throughput and lower packet loss rate have been achieved. This enables users worldwide to access the Super AI platform more quickly and stably, greatly enhancing the user experience.