Summary of Key Points
Yuanmu Intelligence is an industrial AI company founded by a team with expertise in AI. Its founder, Jiang Changhao, decided to leave the financial technology sector, where “smart money” was easily made, and focus on the often overlooked market of small and medium-sized factories in the manufacturing industry. Instead of following the traditional approach of “digitizing first and then intelligentizing,” Yuanmu uses AI agents to solve complex issues such as production scheduling and process optimization. Their core product, “Today’s Scheduling,” helps factories save 90% on scheduling time and increase equipment utilization by 15%. With affordable pricing (ranging from 50,000 to 100,000 yuan per year), the company has already partnered with dozens of machining factories in various high-end industries. Their goal is to improve the overall efficiency of factories throughout the value chain, shorten delivery times, and pave a new path for Chinese manufacturing from relying on physical strength to leveraging intellectual prowess.
Why Choose the Challenging Industrial Sector Instead of Financial Technology?
Jiang Changhao’s previous two startups were in the financial technology field, where data was well-organized and customers were willing to pay—typical cases of “smart people making smart money.” But why did he switch to industry?
- Challenges in Finance: The financial sector faces increased regulation, and there is a high level of homogeneity among thousands of banking companies. With the emergence of large models, it’s becoming increasingly difficult for startups to gain a competitive advantage (since one successful company’s model can be easily replicated by others).
- Opportunities in Industry: Industrial scenarios are diverse, and large models can handle unstructured data like documents and spreadsheets, which addresses issues that traditional digitalization methods fail to address (such as the expertise of experienced workers and process reasoning). In 2024, after visiting over a hundred factories, Jiang realized that Chinese factories could directly adopt AI without going through the lengthy digitization process first—this presents a significant opportunity for breakthrough development.
Breaking Conventions: Can AI Help Factories Work Without First Going Digital?
Traditional industrial software follows a step-by-step approach: install sensors, collect data, build databases, and then introduce intelligence. However, Yuanmu takes a different route:
- Traps of Traditional Digitalization: The data collected in the past has been easy to gather but not necessarily useful (e.g., equipment operation times). Jiang’s team found that maintenance documents often lack valuable “hidden knowledge” about how experienced workers repair machines. This unstructured information is useless without proper processing.
- Yuanmu’s Gradual Approach: Instead of trying to create a perfect AI agent from the start, they focus on developing tools that workers will actually use. For example, they added a recording function to maintenance tools, allowing AI to gradually learn from workers’ experiences through interaction. This approach is more practical for an AI team without prior industrial experience.
Scheduling Challenges in Small and Medium-Sized Factories: “Today’s Scheduling” Saves Time
Production scheduling is a major issue for factories. Large companies invest millions in advanced scheduling systems (APS), but small and medium-sized factories can’t afford them and don’t have complete data sets, so schedulers rely on Excel for manual adjustments, which often leads to disruptions. Yuanmu’s “Today’s Scheduling” solves this problem:
- Product Logic: The AI provides initial suggestions, and schedulers can then refine these using a Gantt chart, with real-time recalculation of any conflicts.
- Visible Results: Time savings of 90%, equipment utilization increased by 15%, and delivery times reduced by 15% (proven in tests).
- Affordable Pricing: The service costs between 50,000 to 100,000 yuan per year, with the price varying based on data usage. For factories with expensive CNC machines, a 15% increase in equipment utilization is a significant benefit.
- Avoiding Privatization: To maintain a rapid iteration pace of two updates per week, they declined requests from large customers for customized deployments to avoid maintenance complexities.
Beyond Scheduling: AI Optimizes Entire Factory Processes
“Today’s Scheduling” is just the beginning. Yuanmu plans to deploy AI across the entire factory value chain:
- Pre-Sales Quotings: AI analyzes unstructured customer requirements (drawings, emails) to automatically calculate material costs, procedures, and delivery times, reducing order processing time.
- R&D Design: AI helps engineers convert 3D designs into 2D drawings, saving 50% in repetitive work.
- Process Optimization: Multi-modal models are used to interpret part drawings and generate process lists for scheduling (product name: “Drawing Simplifier”).
- Data Feedback Loop: Each adjustment by schedulers provides more data for the AI to learn from, making it more efficient over time. The ultimate goal is to shorten the entire cycle from order receipt to delivery—just as Foxconn was able to secure Apple orders because of its faster response times.
How Does Yuanmu Compete Against Giants and Established Players?
Yuanmu faces competition from giants like Siemens/Huawei, traditional MES/ERP vendors, and other AI startups. But Jiang is confident:
- Differentiated Competition: While giants focus on serving large clients or using traditional tools, Yuanmu focuses on “system-level” optimizations.
- Comprehensive Skills: Industrial AI requires a combination of algorithms and engineering expertise (e.g., using cost-effective models for different scenarios).
- Data Advantage: Each industry’s factory data is unique, and Yuanmu has accumulated valuable data from small and medium-sized factories that others cannot access.
Jiang Changhao believes: “If you want to build a great company, don’t fear competition.” He sees the potential in the untapped efficiency of Chinese factories, which could become a new competitive advantage for Chinese manufacturing.
(The entire text is written in plain language, making it easy for non-financial professionals to understand Yuanmu Intelligence’s business model and value.)