Introduction

Deep learning is the fastest-growing field of artificial intelligence, helping computers understand data such as images, sounds, and text. Now through layers of neural networks, computers can observe, learn, and react to complex situations just as well as humans, or even better.

The GISAI-artificial intelligence computing platform developed by our company is a deep learning model developed for data scientists. It can be placed on a platform to automatically generate a Web API for direct calls to all services that require this service. This approach is different from the current public cloud GCP, AWS. AI services are similar

However, this platform can help you to have your own AI platform. You can purchase the required GPU/CPU hardware environment by yourself. After installing the GISAI platform, multiple AI services can be integrated into the same hardware environment, which can greatly save the cost of renting public cloud.

Application field

When your company or department needs customized AI services or its own AI algorithms need to be placed in its own private cloud, you can purchase this software to achieve self-built AI cloud capabilities.

Usage scenario

The image recognition services provided by our company are placed on the GISAI platform for publishing WEB API.

Features

  • Through the docker/Container technology, multiple groups of different WEB API services can be realized at the same time
  • Multiple groups of AI algorithms can share the same GPU resources
  • Provide a jupyter editor so that data scientists can write algorithms and test AI services in real time through nodejs and python languages to achieve rapid development and deployment capabilities
  • Provide Git support, so that large data and algorithms can also have version management requirements
  • The ability to support a large number of simultaneous calls in the WEB API provides direct online replication of multiple AI services, so that the services of large numbers of calls can operate smoothly
  • Provide each docker performance monitoring service

Contacts

04-24516609#260

Jeffrey