A Conversation with Tang Chunfeng, CEO of Super Engine Digital Intelligence: Competition in the AI industry is shifting its focus from model training to inference deployment, and full-stack solutions are accelerating the rollout of enterprise intelligent applications.
On July 17, the 2026 World Artificial Intelligence Conference opened in Shanghai. As an important content window for 36Kr to go deep into the WAIC venue for the third consecutive year, the "Kr-Talks Future" live broadcast room also launched on-site dialogues simultaneously on the first day of the conference. Tang Chunfeng, Founder and CEO of SuperEngine AI, accepted an exclusive interview with 36Kr's "Kr-Talks Future" at the WAIC venue. Focusing on topics such as the implementation of AI inference applications, full-stack solutions integrating software and hardware, and industry application practices, he shared SuperEngine AI's overall layout from software and hardware to industry deployment, as well as how to improve AI application efficiency and reduce enterprise implementation costs through technical solutions.
The theme of this year's WAIC is "Smart Partners, Co-Create the Future". As the artificial intelligence industry enters the stage of deepening applications, the industry's focus is gradually shifting from competition in model capabilities to the creation of application value. Compared with the discussions around parameter scale, model performance and technological breakthroughs in the past few years, the 2026 AI industry is paying more attention to a core issue: how AI can truly enter enterprise business processes, how to lower application thresholds, and how to create value through actual business results.
Under this trend, artificial intelligence inference applications are becoming an important direction for industrial implementation. For enterprises, AI applications not only require model capabilities as a foundation, but also efficient and stable software and hardware support to reduce deployment costs and improve operational efficiency. How to transform AI capabilities into scalable and replicable industry solutions has become the key to promoting the further implementation of artificial intelligence.
SuperEngine AI is continuously exploring in this direction. By building a full-stack solution for artificial intelligence inference applications, SuperEngine AI combines software and hardware capabilities to improve the implementation efficiency of industry applications. At present, the company has launched cooperation in multiple fields such as computing power services, new drug R&D, financial services, and embodied intelligence, and helps customers improve R&D efficiency and reduce application costs through its artificial intelligence acceleration platform. For the AI industry, the key to moving from technological breakthroughs to business implementation lies in building application capabilities that can truly serve industry needs.
The following is a transcript of the dialogue, edited by 36Kr:
36kr: Hello, Mr. Tang, welcome to our live stream. At this year's WAIC, what is the core content that SuperEngine AI most wants to showcase? In other words, what is the highlight that you most want the outside world to remember this time?
Tang Chunfeng: The development of artificial intelligence has actually rapidly shifted from large model training to the implementation of inference applications. From SuperEngine AI's perspective, we have fully seized this key transformation since last year, and built a full-stack solution around the implementation of artificial intelligence inference applications. This solution covers both software and hardware, and improves the implementation efficiency of industry applications through the combination of software and hardware. This is the key direction that SuperEngine AI is currently focusing on promoting.
36kr: What aspects of customer problems does the solution demonstrated this time mainly solve? Is it more focused on efficiency improvement, cost optimization, stability guarantee, security capabilities, or business experience enhancement?
Tang Chunfeng: This is a very important link in the development of artificial intelligence, which is moving from model training to productivity improvement. The current industry is also paying attention to "AI Factories" and Tokenomics, which essentially means how to get more output with less input. Therefore, our core goal is to effectively improve efficiency and reduce costs through technical solutions, helping enterprises realize the large-scale implementation of artificial intelligence applications.
36kr: Are there any typical scenarios that have been implemented or are currently in pilot operation that you can share? What are the issues that customers are most concerned about during the deployment process, or what is the biggest challenge?
Tang Chunfeng: At present, there are mainly two types of demands for artificial intelligence implementation. One type is computing power-related enterprises, which pay more attention to improving computing power efficiency. We have cooperated with a large number of computing power companies in China. The other type is private deployment scenarios, such as new drug R&D, financial services, embodied intelligence and other fields, which have high requirements for data security and business adaptation.
Over the past year or more, we have achieved some implementation results in these fields. For example, in Yizhuang, Beijing, we cooperated with Yikang Pharmaceutical, a pharmaceutical company focusing on cancer new drug R&D. By delivering the full-stack solution of the artificial intelligence acceleration platform, we effectively improved the efficiency of new drug R&D and reduced R&D costs at the same time. At present, this type of solution has enabled multiple projects to be implemented in the field of new drug R&D.
36kr: In the next 6 to 12 months, what directions does Mr. Tang most hope to make breakthroughs in? In other words, after WAIC, what actions of SuperEngine AI can the outside world focus on?
Tang Chunfeng: In the next 6 to 12 months, I believe that artificial intelligence inference applications will further accelerate their development, especially the demand around "AI Factories" will see explosive growth. We hope to quickly build and improve the full-stack solution through the R&D and testing platform, improve the productization efficiency of the solution, and better support the large-scale implementation of artificial intelligence inference applications. In the future, as this field enters a stage of rapid development, SuperEngine AI also hopes to continuously promote industry application innovation and industrial development.