It is neither an image recognition tool nor a product search service: Yizhao aims to build "buyer-centric e-commerce" in the AI era, a model that first identifies products, then generates appropriate selection options, and finally facilitates the completion of purchase.
Consumers are facing an increasingly wide range of products, but the real difficulty is often not "failing to find the goods", but struggling to figure out what the product itself is, whether the merchant's claims are valid, and whether the price is reasonable. To address this issue, Shenzhen Yingquan Artificial Intelligence Co., Ltd. is developing a consumer decision-making product called "Yizhao". The project starts with wine, and provides consumers with product identification, attribute judgment, price reference and risk warning through AI recognition, public information acquisition and industry rule judgment. Launched for about two months, the project currently has a total of about 5,000 users, mainly from organic traffic on short video platforms, and about 40% of the users have completed a second query. The project is now seeking RMB 5 million in Pre-A financing.
First solve the problem of "not understanding the product"
Traditional e-commerce has greatly improved product supply and transaction efficiency, but product information has also become more complex. For ordinary consumers, the same category of products may have multiple layers of information at the same time, including brand stories, origin, grade, specifications, channels, promotions and streamer pitches, making it difficult for consumers to determine which information is truly relevant to product value in a short period of time.
This problem is more prominent in categories with high information asymmetry such as alcohol, health supplements, cosmetics, food and luxury goods. Consumers can usually see the introductions provided by merchants, but lack independent professional judgment capabilities. Search engines and general large language models can help obtain information, but when it comes to judging specific products, it is still necessary to combine industry rules, price systems, product standards and empirical knowledge.
Therefore, the product form of "Yizhao" is neither a traditional product search tool nor a simple image recognition tool. Taking wine as an example, after the user takes a photo of the wine label, the system will not only identify the brand, producing area and wine style information, but further determine what type the product belongs to, whether the relevant publicity complies with regulations or industry facts, what the usual price range of similar products is, and explain the judgment basis to the user. The project aims to convert the judgment capabilities that used to be mastered only by industry practitioners into tools that ordinary consumers can use directly.
Starting from wine, turning industry judgment into systematic capability
"Yizhao" currently selects wine as the verification category first, not because of the alcohol market itself, but because the founding team has long-term experience in this industry, which can verify the accuracy of AI output and disassemble implicit industry experience into clear rules.
The technical path of the product does not only rely on the large model to generate answers directly. AI is mainly responsible for image recognition, information extraction, data sorting and semantic understanding, while industry knowledge, expert rules and judgment logic are responsible for forming the final conclusion. For problems that cannot be confirmed only by public data, the system will retain uncertainty instead of forcibly giving an absolute judgment. The project hopes to gradually precipitate professional judgment into a replicable systematic capability by continuously adding cases, rules, expert knowledge and manual review.
This approach also determines the future expansion mode of "Yizhao". The project will not simply copy the wine recognition function to more categories, but give priority to entering fields where consumers have high judgment costs and large product information gaps, and then establish corresponding knowledge, rules and verification systems for different categories. The long-term goal is to form a cross-category basic consumer judgment capability.
The founder has long been engaged in wine business, supply chain, sales management, instant retail and live streaming e-commerce in the past. He once set up 58 instant retail stores for alcohol in Shenzhen, and participated in multiple links including product procurement, channels, pricing and content-based user acquisition. The product design, industry rule sorting, content user acquisition and early operation of the current project are mainly promoted by the founder. About two months after launch, Yizhao has accumulated about 5,000 users without large-scale paid traffic investment. Existing users mainly come from organic short video traffic, of which about 40% have completed a second product query. This data is currently regarded by the team more as an early signal of whether users are willing to reuse the behavior of "making judgments before purchase", rather than a commercialization result.
From judgment to e-commerce entry, seeking RMB 5 million financing in this round
The long-term direction of "Yizhao" is not limited to being a product identification tool, but hopes to enter the e-commerce transaction chain from the "judgment" link. As AI gradually participates in search, comparison, recommendation and shopping decision-making, the e-commerce entry may further move forward from the traditional product shelf to the moment when consumers put forward their demands. Compared with entering the mall first and then facing a large number of products, the path envisioned by Yizhao is that consumers first tell AI what they want to buy, what their budget is, and what scenario the purchase is for, then the system completes product judgment, screening and comparison, and finally leads to the transaction.
The project defines this direction as "buyer-side e-commerce". Its core is not to add a new product mall, but to place the consumer's judgment link before the transaction: judge the product first, then form a choice, and finally make the purchase. For Yizhao, the most important task at this stage is still to improve the accuracy and stability of professional judgment, so the project is in no hurry to launch product recommendations on a large scale. The team believes that if the judgment capability itself has not been established, entering recommendation and transaction prematurely will easily make the product degenerate into an ordinary shopping guide tool.
In terms of business model, Yizhao plans to keep basic queries free of charge. After entering the transaction stage in the future, the platform plans to take fixed transaction service fees as the main source of income, which is currently envisioned to be about 3% of the transaction amount, and try to avoid relying on advertising or paid ranking to affect product judgment. The target users are first ordinary consumers who buy products with high information asymmetry and want to reduce decision-making costs. With the expansion of categories, the market boundary will also expand from wine to more consumer goods fields that require professional judgment.
In this financing round, the project plans to raise RMB 5 million with 15% equity to be transferred. The funds will be mainly used for the construction of product and engineering teams in the next 18 to 24 months, improvement of product data and expert rule systems, verification of judgment accuracy, and expansion of more categories. The team currently hopes to prove three things first: whether consumers will form the usage habit of making judgments before purchase; whether professional judgment capabilities can be stably replicated to the second and third categories; and whether this entry can be further naturally extended to screening, recommendation and transaction.
If the above path is verified, "Yizhao" hopes to become a consumer-facing product judgment entry in the AI era, and further explore the e-commerce form from "helping consumers understand products" to "helping consumers complete purchase decisions".