Wangsu Technology launches edge AI gateway to help developers build AI

On July 11, Wangsu Technology announced the launch of a new AI service - Edge AI Gateway, which provides enterprise users with a secure, reliable, high-performance, and easy-to-observe integrated model service management solution to help developers in their AI building journey .

Zeng Zhiquan, product development director of Wangsu Technology's CDN, said that AI has become the focus of a new round of innovation. Many companies want to use AI but lack relevant technology or experience. Wangsu is the first in the industry to launch an edge AI gateway, hoping to help companies apply AI better, simpler and more securely.

It is reported that Wangsu Edge AI Gateway aims to build a bridge between model services and AI applications for enterprises . Whether individual developers or enterprises, they can use Wangsu Edge AI Gateway to more easily manage and integrate model services, thereby better carrying out AI application innovation.

Large model services face management, cost and security challenges

In recent years, the emergence of the GPT big model has set off a big model competition around the world, and as a result, AI applications in various industries have accelerated. Today, in application development, integrating big models has become a general trend to improve application experience and functional depth.

In this context, a new service model, MaaS (Model as a Service), has emerged in the industry. Model providers rent out large model capabilities to third-party companies through API interfaces, allowing companies to interact with them through HTTP requests or other means. Service providers are responsible for updating and maintaining the models, and companies can directly enjoy the latest model improvements.

However, when enterprises use multiple models or providers, they face high AI application costs. In order to use the model services of various AI service providers at the best cost , enterprises face management and operation and maintenance challenges. How to better control the use of models, reduce costs, and clearly understand the current status of AI use and service issues have become the inherent demands of various enterprises in building AI applications .

In addition, it should be noted that in the process of using model services, enterprises also face risks such as sensitive data leakage and content non-compliance. At present, most corporate organizations have not yet started to address AI -related risks , or are in the process of addressing them.

"In response to the above challenges, the industry urgently needs a solution that can reduce the difficulty of managing and expanding model services and improve AI availability and security. Wangsu Edge AI Gateway can meet this demand, allowing enterprises and developers to focus on building AI applications rather than managing complex integrations," said Zeng Zhiquan .

Wangsu Edge AI Gateway , manageable, controllable, and observable multi-model services

Helping enterprises grasp innovation in technology applications has always been Wangsu Technology’s mission.

Zeng Zhiquan said that in the successive iterations of Internet applications, Wangsu Technology has provided important support for the large-scale implementation of various emerging applications and provided global enterprises with one-stop, efficient and secure service guarantees. Our accumulated complete technical capabilities and profound scenario experience can be reused in the current wave of AI innovation , and we can start in the AI ​​era together with enterprises .

Based on insights into enterprise needs, Wangsu Technology was the first in the industry to create the Wangsu Edge AI Gateway , creating a new service paradigm in the industry.

Specifically, in general, enterprises need to do some integration development in advance to manage the multiple large models they access. Wangsu Edge AI Gateway provides a universal, integrated solution. Enterprises can create edge AI gateways and add channels through a unified control platform to achieve docking with multiple AI service provider models. The platform supports the manageability, controllability, and observability of multi-model services.

Among them, at the access level, Wangsu Edge AI Gateway has completed the integration with multiple mainstream AI service providers in advance . Enterprise developers do not need to modify the code frequently. They only need to configure the gateway in the console. With one access, they can use multiple model services, greatly improving access efficiency.

At the management level, Wangsu Edge AI Gateway supports special configurations. Enterprises can flexibly define priority, load balancing, speed limit and other policies on the AI ​​gateway, configure different model services on demand, achieve cost control or reduce abuse, and achieve seamless switching in the event of request failure, improving the reliability of AI services . In addition, enterprises can also enable the cache function. Once a request hits the AI ​​cache , it does not need to return to the model provider, reducing the consumption of the model provider tokens, reducing costs and improving response speed.

At the operational level, Wangsu Edge AI Gateway provides data visualization capabilities, aggregates indicators from multiple model providers, supports analysis and display of the number of requests, number of tokens, number of caches, number of errors, etc., and provides real-time logging, allowing enterprises to observe and control the use of model services to optimize model service strategies.

Wangsu Edge AI Gateway , security and performance guarantee for multi-model services

Zeng Zhiquan further introduced that while achieving efficient access and management of multi-model services, the core value of Wangsu's edge AI gateway lies in ensuring the security, high performance, and customizability of model services.

Security is critical to AI applications . In terms of security, enterprises can configure prompt word filtering rules or perform content security audits on Wangsu edge AI gateways to ensure that interactive content is safe, compliant, and respectful of users .

In terms of service performance, relying on Wangsu's powerful global resource network, users can access the edge distributed AI gateway nearby to ensure the stability and speed of access and reduce the impact of abnormal factors such as cross-border network fluctuations on user experience.

It is worth mentioning that Wangsu can support developers to achieve a certain degree of serverless development at the edge location closest to users. By combining Wangsu's edge AI gateway and Wangsu's edge function , some model calls or intermediate processing logic can be flexibly customized, and services can be launched within 1 hour, greatly saving R&D personnel time.

It is reported that at present, Wangsu's edge AI gateway has opened a limited-time free test to support users' AI experience . In addition, Wangsu has deepened its enterprise applications and focused on specific AI scenarios . It has also created edge vector databases, edge model inference services, etc., and plans to open them to the public in the next step.

At that time, enterprises can easily integrate public or proprietary models, reasonably manage model applications through the edge AI gateway, obtain multi-model operation data, or combine with edge functions, edge KV storage, application engines, etc., flexibly build and expand edge AI applications , and perform AI tasks such as intelligent question and answer, content creation, retrieval enhancement, and image recognition .

With the development of the AI ​​era , Wangsu Technology's capabilities in the AI ​​field will continue to deepen. Zeng Zhiquan said that Wangsu Technology will continue to build a more agile, adaptable and secure AI ecosystem, so that more companies can embrace AI and gain innovation and growth from it .