Ping An Technology: Leveraging AI+Cloud to Enable a Better Future for Businesses

In the past few years, artificial intelligence (AI) and cognitive technologies have been changing the way organizations across various industries operate. From utilizing deep learning to processing unstructured data such as images, audio, and text to leveraging multi-modal computing to integrate vision, speech, semantics, and emotions, businesses have been exploring advanced technologies to improve business outcomes. Considering these market dynamics, China-based Ping An Technology employs AI, cloud and other cutting-edge technologies to develop and operate mission-critical platforms and services to support businesses in financial, medical, automotive, real estate, and smart cities segments.

Ping An Technology is the technology arm of Ping An Insurance Group, the largest insurer in China who strives to be a world-leading technology-powered retail financial services group. Ping An Technology builds cognitive technologies and incubates an ecosystem of solution providers. It follows its corporate philosophy of “technology changes lives” with “Cloud Infinity” as a value proposition, under the leadership of Ericson Chan, CEO of Ping An Technology. Another major solution of Ping An Technology is ‘Ping An Cloud,’ a cloud offering with its own Industrial Data Center infrastructure that can provide full-stack cloud service including IaaS (Infrastructure as a Service), PaaS (Platform as a Service), and SaaS (Software as a Service) layers. This suite of cloud solutions caters to over 500 million existing users and enables a better future for its customers in China and overseas.

The other widely used offerings of Ping An Technology is its AI and cognition capabilities which includes smart medical imaging, speech recognition, and multi-modal authentication solutions.

These solutions provide significant benefits for businesses operating in the financial and healthcare sectors. The company has also introduced a robotic process automation tool (RPA), An Xiao Feng, for insurance, banking, and securities sectors. The process automation robot aims at reducing labor cost and enhancing the customer experience by replacing manual operations with the simulated human operation. In addition to the RPA solution, our expertise in cognitive technologies also includes a full-scale robotic customer service system for retail enterprises. This system consists of online text-based customer service robots, telephone customer service, agent assistant, and physical store service robot.

The company’s capabilities, however, go far beyond these solutions as it has many other intelligent offerings in its portfolio. As an example, we provide a comprehensive ‘personal identification service platform,’ which integrates several cognitive technologies such as facial, voiceprint, OCR, text, and image recognition. The platform assists enterprises in delivering government-authorized information checking services ranging from citizen identification to banking and telecommunication service registration. “Our extensive range of intelligent solutions and services impacts the three primary abilities. First, to identify faces through micro-expression recognition, and vehicle damage assessment; second, to hear with voiceprint recognition and automatic speech recognition and third, to read, write and talk through NLP/NLU, chatbot, and text to speech,” adds Dr. Xiao Jing, Chief Scientist of Ping An Technology and Executive Committee Member of Ping An Insurance (Group) of China, Ltd.

With such an extensive portfolio of advanced solutions, Ping An Technology has been addressing various challenges related to several industrial applications over the years. One example in which the company witnessed remarkable results is vehicle damage assessment, a service that has been successfully launched in three property insurance institutions. As the number of property insurance institutions increases, the traditional picture recognition approach does not work well as it causes timeout as the detection of auto parts takes about 200ms to process an image. The Ping An Technology team approached this challenge by using model compression, quantization, pruning, and other techniques to optimize the performance of the recognition algorithm. Consequently, Ping An Technology can analyze a picture in less than 20ms whereas the conventional picture recognition takes more than 200ms to process an image.

Ping An Technology is determined to become a global leader in AI and cloud computing. “We will continue to increase our investment in research while dedicating 1 percent of the revenue of the group company, Ping An Insurance Group, for scientific research with plans to invest 100 billion RMB over the next decade,” concludes Dr. Xiao.