.jpg)
As generative AI is revolutionizing the operational models across industries, how can enterprises leverage this new technology to enhance production efficiency and make AI their indispensable ally in daily work? Netron Information Technology, in collaboration with TechOrange, co-hosted the workshop "Navigating New AI Application Scenarios: Unlocking Enterprise Innovation Guidelines and Opening New High-Efficiency Production Models" on November 19, where they explored innovative and efficient new work models with the attendees.
"The difference between AIGC (Artificial Intelligence Generated Content) and discriminative AI lies in the ability to learn from vast amounts of data and generate diverse outputs based on needs, autonomously creating new text, images, and videos," said Chen Zhen, Senior Business Manager at Netron Information Technology. "The core technology of AIGC is the Transformer architecture and Generative Adversarial Networks (GANs). GANs consist of a generator and a discriminator. The generator creates fake data for the discriminator to identify, and through this 'adversarial' process, the data generated by the generator becomes more realistic."
.jpg)
Netron Information Technology Senior Account Manager, Chen Zhen, delves into the topic "Embracing the AIGC 2.0 Era: Building a Cloud Service Model that Combines Flexibility and Security."
Chen Zhen mentioned that AIGC technology has diverse application scenarios. For example, in the media and entertainment industry, AIGC can assist in generating presentation titles and summaries to improve work efficiency. In film production and game development, AIGC can generate sketches based on demand, allowing users to only make subsequent corrections, thus shortening the creative cycle. In e-commerce and marketing, AIGC can create compelling copy based on current consumer habits and needs. In market analysis, AIGC can assist users in formulating future sales strategies. In education and training, AIGC can generate customized learning content based on students' study materials and progress, becoming a virtual teaching assistant.
With the upgrading of AIGC technology and the optimization of models, more diverse content can be generated by integrating data, but this also raises concerns for businesses, such as providing outdated information, data privacy and protection issues, and the substantial costs associated with system development, deployment, and maintenance. Chen Zhen emphasized that Netron Information Technology's NAVI Enterprise AI Knowledge Management System can overcome these three challenges: "Firstly, all data in the system is stored in the client's own AWS account and will not be used for other purposes. It also ensures compliance, addressing information security issues. Regarding cost, the system manages token quotas to help businesses easily control their budgets. It also supports knowledge base sources like AWS S3 and websites, providing uninterrupted 24/7 public cloud maintenance services, saving on development costs. As for data accuracy, NAVI only retrieves, organizes, and answers based on uploaded data, preventing AI hallucinations and incorrect answers."
The application scenarios of the NAVI Enterprise AI Knowledge Management System in the financial and media industries.
"NAVI is an AI product specifically designed for enterprises, offering external integration services," explained Xu Kaijun, Project Manager at Netron Information Technology. He shared the product roadmap for the NAVI Enterprise AI Knowledge Management System. In customer service applications, it is already capable of assessing different input requirements and providing assistance, even integrating with the company's front-end pages. For educational and training applications, the NAVI system can analyze video content, and by uploading files to the knowledge base, it enables learning through natural language Q&A, saving more manpower costs for the enterprise. Xu further mentioned that if an enterprise has its own front-end or uses the LINE app, it can connect the AI system to the front-end page via the Open API provided by NAVI, enabling intelligent services without the need for additional AI system development. "Typically, the entire AI system can be deployed in one to two days, and once internal testing is completed, it can be ready for customer use."
.jpg)
Xu Kaijun, Project Manager at Netron Information Technology, stated that the NAVI Enterprise AI Knowledge Management System helps enterprises achieve tangible benefits, such as freeing up human resources and protecting data privacy and security.
Xu Kaijun further shared the application scenarios of the NAVI Enterprise AI Knowledge Management System. Netron Information Technology leverages Amazon Bedrock (fully managed service) and Amazon Kendra (intelligent enterprise search service) to build a generative AI solution. This solution helps securities regulatory agencies quickly analyze and summarize financial reports and earnings call videos, enhancing overall work efficiency and reducing errors in manual reviews. The second case involves a traditional newspaper company, where the NAVI Enterprise AI Knowledge Management System helps organize a large historical news database and uses AWS cloud storage and archive management technology to improve news search efficiency while optimizing press release generation. Xu Kaijun emphasized, "More applications depend on the user's imagination. As long as you can imagine it, AI applications are not that limited."
Xu Kaijun summarized the pain points that the NAVI Enterprise AI Knowledge Management System solves for businesses, including "freeing up human resources." Through AI assistants, content generation, and data analysis, it greatly saves time and costs. Additionally, "integration and scalability" is a key feature, with Netron Information Technology offering open APIs that can integrate with other application systems or services, and expand support for large language models like Azure OpenAI and AWS Claude3. Lastly, "privacy and security" is a top priority, with NAVI integrating Windows AD and Azure AD identity authentication mechanisms, allowing different knowledge bases to be set up according to data or department. "We are continuously improving product performance and plan to develop support for oral communication applications in Q1 next year. We also hope to build a knowledge management ecosystem to help customers quickly implement AI solutions and deepen cooperation with service products," Xu Kaijun added.
