Shareholders' share locking commitments. In December 2022, the company issued a share lockdown announcement. At that time, 75% of shareholders promised to extend the lockdown. In January 2024, the company continued to issue share lockdown announcements, and nearly 60% of the original shareholders promised to continue extending the lockdown period. After the past year, we expect that some shareholders have chosen to withdraw through the secondary market or other methods other than the secondary market, so we carefully estimate that after this commitment, about 10% of the shares may still be willing to reduce their holdings. Judging from the performance of the secondary market, the reduction in old stock holdings in the past six months has indeed had a big impact on the company's stock price.

About company fundamentals. We believe there have been no significant changes in the company's fundamentals so far. Since its listing, the company's revenue has continued to grow rapidly, especially in the AI+ manufacturing industry. As of the first half of 2023, the company's total revenue continued to grow at a rate of 43%, while the AI+ manufacturing industry grew at a rate of 81%. However, ToB projects are often forward-looking, including manufacturing orders, that is, some 2022 projects will only be officially completed in the first half of 2023. In other words, the performance in the first half of 2023 may reflect more of the 2022 projects. Considering the factors of the major economic cycle, we took the initiative to lower the company's revenue forecast for the full year of 2023 in September 2023, and the company stopped secondary market repurchases in the second half of 2023, which indicates that the company has actively made adjustments based on cash flow safety. Similarly, we believe that it should also include the company's signs of project expansion and contraction.

Judging from the company's project system, it includes a combination of software and hardware projects and pure software projects. Projects combining software and hardware mainly include intelligent visual solutions, such as iron and water transportation, quality inspection, intelligent robotic arm factories, etc., while pure soft projects are industrial intelligent execution systems with MOM as the core. Software and hardware integration projects are generally large in scale, have a long cycle, and low gross margin. Pure software systems are generally relatively small in scale and short in cycle, but the gross margin is high. Therefore, if you are pursuing high revenue growth, a soft and hard combination project is a better choice, while if you pursue safe cash flow or achieve break-even as soon as the company's management has always emphasized, then a pure soft project seems to be a better choice. We believe that there is a balance in this, and it is also a test for management decisions, especially when the economic cycle is weak. The penetration rate of China's intelligence in the manufacturing sector is still relatively low, and medium- to long-term market demand is not a problem.

About the big model. The company's practical implementation based on the 15B parameter model has had initial results. Currently, it is more at the application level of certain segments. The company is also planning to launch a large industrial vertical model with larger parameters and develop it in depth. According to our understanding, in the process of informatization and digitalization, large manufacturing enterprises often have hundreds of application systems within them. In the intelligent era, how to simplify and integrate many application systems, and perform intelligent interactive analysis to truly achieve the goal of intelligent intelligent manufacturing, cost reduction and efficiency, is both a direction and a real requirement, especially in the era of large models. We expect that in the next 3 to 5 years, in the process of restructuring the enterprise intelligence system, the big model may become the underlying technical architecture platform as infrastructure, fully integrated into all aspects of the enterprise's daily operation, manufacturing, etc., thus forming an effective technical barrier and industry moat.

Risks: Liquidity improvements fall short of expectations and stock prices are under pressure due to the withdrawal of old stocks; revenue decelerates due to the company's business line adjustments and contractions; progress in implementing the big model falls short of expectations;