Chexin Intelligent's self-developed AI Agent solves the user operation challenges in the automotive industry.
The project "CarMind AI" under Hangzhou Zhuimi EV Research Automotive Technology Co., Ltd. launched a RMB 3 million seed round financing on October 30, 2026. The funds raised in this round will be invested in R&D of AI Agent and platform products, business expansion and car owner operation, and construction of aftermarket supply chain. The phased goal is to realize the paid conversion of car owner organizations, land 10 benchmark car owner organizations and 2 auto enterprise cooperation cases, and complete cross-brand product replication and commercial implementation.
In the era of stock vehicles, car owner user operation is facing multiple bottlenecks
The domestic automotive market has fully entered the stock competition stage, with the industry focus shifting from new vehicle sales to long-term car owner relationship operation. As of the end of June 2026, the number of motor vehicles in China reached 371 million, among which new energy vehicles amounted to 48.97 million; the global automotive aftermarket market size exceeds 1 trillion US dollars, and the expected size of China's automotive maintenance and repair market in 2025 can reach 1.74 trillion yuan. As the proportion of vehicles with an age of more than 7 years breaks through 50% and the proportion of newly registered new energy vehicles in the first half of 2026 is close to 50%, the long-term consumption value of car owners in maintenance, modification, insurance, vehicle replacement and other scenarios continues to be released.
However, current car owner communities generally have four structural pain points: vague commercialization path, operation highly dependent on the personal energy of community leaders, high difficulty in user retention, and community growth bound to core managers. Once the leader's energy declines, the community activity and conversion capacity will drop rapidly. Existing players in the track all have shortcomings: self-developed apps of auto companies only serve a single brand and cannot support cross-brand car owner organizations; general SaaS tools lack capabilities in automotive vertical scenarios, and do not understand native scenarios such as car owner activities, maintenance, color changing, and car owner gatherings; the monetization of automotive content platforms is mainly based on advertising, which is difficult to support the conversion from community private domain traffic to local services.
In addition, car owners themselves have the demand for owning multiple vehicles and joining multiple cross-brand car owner organizations, and existing products are difficult to adapt to the flexible multi-community identity system, which is also a problem that has been ignored by the industry for a long time.
Four-layer foundation + vertical AI Agent to build an intelligent network for car owners
Positioned as the intelligent infrastructure for user relations in the automotive industry, CarMind AI takes the core idea of shifting from vehicle intelligence to car owner intelligence and building an intelligent network for Chinese car owners. The product builds a four-layer foundation to form a closed loop: My Car Club, the C-end entry for car owners; CarMind Club OS, the exclusive operation system for car owner organizations; CarMind Agent System composed of four types of automotive vertical professional AI Agents; and the intelligent associated data graph of car owners, to realize the effect of "entry precipitates car owner relations, system undertakes organization operation, Agent releases automatic operation capability, and data empowers aftermarket business".
The project's AI Agent is not designed to directly replace the decision-making of community leaders, but to position itself as an operation assistant: it automatically analyzes car owner dialogues, identifies real demands such as maintenance, color changing, insurance claim, group travel, price comparison, outputs operation plans, activity planning, and service recommendation suggestions, and the final decision-making right remains in the hands of car owner managers, which balances efficiency and does not damage the original trust ecosystem of car owner communities. Meanwhile, all user data is processed through desensitization and aggregation, with strict data compliance to avoid private domain data risks.
Target users are divided into three categories: cross-brand car owner leaders and car owner organizations, user operation departments of auto enterprises, and offline automotive aftermarket stores (film pasting, maintenance and modification stores). In terms of business model, at the present stage, the core paths are SaaS service fee for car owner organizations, AI value-added services, and commission sharing of aftermarket service transactions. Relying on the trust network of car owner leaders to acquire customers at low cost and convert community traffic into offline in-store consumption, the project has verified the commercial closed loop through model stores: the monthly revenue of the model offline service store exceeds 300,000 yuan, and the conversion link from community traffic to automotive services has been realized.
Nearly 24,000 registered car owners available, serial entrepreneur rooted in the front line of communities
At present, the platform has accumulated resources of 97 car owner organizations, covering 57 cities, 15 automotive brands, 24 provincial-level regions, with 24,000 registered car owners and about 1,000 daily active users; the AI in the platform can analyze more than 200 real demands of car owners every day. Online commercialization has not been fully launched yet, and the current focus is on user accumulation and product polishing. The core observation indicators include organization activity rate, 30-day user retention rate, organization paid conversion rate, AI intention confirmation rate, and customer renewal rate.
The founder & CEO graduated from University of Electronic Science and Technology of China, with 12 years of experience in automotive industry product and user operation. He once worked in Alibaba and Alipay, is a serial entrepreneur, and is also the leader of the 10,000-person scale Xiaomi Car Owner Club. He has been rooted in the front line of car owner communities for a long time, and is familiar with the demands of the three parties: community leaders, car owners and stores. The core members of the team cover automotive aftermarket operation, AI Agent R&D, and large-scale internet platform architecture, with both front-line practical experience in communities and engineering implementation capabilities of large models.
"Over the past decade, the industry has focused on vehicle intelligence, but the core factors that truly determine the repurchase of aftermarket services and user referrals are never the vehicles themselves, but the continuous demands of every real car owner. What we are doing is to build this ignored intelligent network for car owners," Tian Geng told 36Kr.
Under the long-term trend of stock vehicle operation, the project plans to rely on the distribution advantage of trust-based car owner organizations and automotive vertical AI capabilities to continuously improve the full-link car owner service system, and finally become an intelligent hub connecting car owners, car owner organizations, auto enterprises and aftermarket service providers.