A former ByteDance product designer starts a business in the "AI pet" field and secures another tens of millions of yuan in financing.

Text by | Wang Fangyu

Edited by | Su Jianxun

To a large extent, the choice of what kind of products a company makes is an extension of the founder's individual emotions and cognition. The AI companion robot company "Ropet Mengyou Intelligence" fully reflects this conclusion.

He Jiabin, the founder of Ropet, was born in 1992 and is married. His career has spanned Microsoft, Baidu, and ByteDance, and he has raised a significant amount of capital for his startup. From a worldly perspective, he has a happy family, a successful career, and many friends.

However, beneath this seemingly successful persona, He Jiabin often feels lonely. He describes loneliness as a certain "weak moment" for adults. This kind of loneliness doesn't stem from being alone but rather from the numbness that arises under social pressure when constantly seeking recognition and success.

How does He Jiabin combat loneliness? He has his own ways. After getting off work at 11 p.m., he feeds stray cats at the subway station. He also specially buys four - wheel drive racing toys for his childhood friends.

During the interview, He Jiabin also brought a Ropet device named "Buddy". When Buddy made a soft "meowing" sound like a kitten during our conversation, He Jiabin gently put his index finger to his lips and made a "shh" gesture, and Buddy quieted down.

He Jiabin, Co - founder/CEO of Mengyou Intelligence       Image source: Company authorization

As the urban life in the concrete jungle makes lonely urbanites increasingly in need of spiritual companionship, He Jiabin believes that this is where the emotional value of "Ropet" can be realized.

He Jiabin told "Intelligent Emergence" a very "counter - intuitive" design concept. While ordinary robots serve humans, Ropet simulates a living being that needs to be taken care of. Through being cared for, users can find comfort.

"We've found that by caring for another 'life', people can gain inner satisfaction and a sense of healing," He Jiabin said.

Another counter - intuitive design is that Ropet doesn't "speak human language".

In the era of the AI large - model boom, making AI toys talk has become a trend. However, after more than a dozen surveys, He Jiabin found that for Ropet's target users, the scenario of chatting with an AI pet for companionship doesn't hold true. In fact, many users are reluctant to talk to robots.

"Companionship doesn't necessarily mean conversation," He Jiabin explained. "Creating a unique and valuable hardware product doesn't rely on the large - model's conversation ability but on how to make it more like a real living being with perception and stress responses."

Based on the above "counter - intuitive" design concepts, the Ropet designed by He Jiabin looks like a kitten. It has a fluffy surface and maintains a temperature of 39 degrees. It can sense human touch and other actions through three built - in sensors and give corresponding facial expressions, sounds, and other feedback.

In early August, Ropet had another successful crowdfunding on the Japanese crowdfunding website Makuake. On the first day of the crowdfunding, it reached a crowdfunding amount of 26.87 million yen (approximately 1.3 million RMB), ranking first on the website's daily list. So far, it has achieved a crowdfunding scale of nearly 50 million yen.

On the Japanese crowdfunding website Makuake, Ropet achieved a crowdfunding amount of 26.87 million yen on the first day.      Image source: Company authorization

According to exclusive information obtained by "Intelligent Emergence", "Ropet Mengyou Intelligence" recently completed tens of millions in Series A1 financing, led by the Beijing Artificial Intelligence Industry Investment Fund, with Fengrui Capital participating in the investment, and Index Capital serving as the exclusive financial advisor.

"Ropet Mengyou Intelligence" was founded by Zhou Yushu, a partner at Zhenzhi Venture Capital, and is committed to creating AI pets for the future. He Jiabin joined in 2023 as the co - founder and CEO.

At this juncture of the new round of financing, we talked with He Jiabin about "how to create an AI companion product for urbanites in the era of the loneliness economy". He has many "counter - consensus" observations in terms of product design and user insight. The following is the edited conversation between "Intelligent Emergence" and He Jiabin:

The loneliness of urbanites: A huge unmet market gap

"Intelligent Emergence": You've mentioned Pop Mart several times in previous interviews. Is there any similarity between Ropet and Pop Mart?

He Jiabin: We're working on the same big proposition, which is solving the problem of loneliness.

Solving the problem of loneliness is a common factor for creating successful products in this era. Pop Mart is the most commercially successful company that solves the loneliness problem through "useless" products. This is something we need to learn from.

"Intelligent Emergence": How do you understand "loneliness"?

He Jiabin: Loneliness is a common state among contemporary people, especially those living in cities. This kind of loneliness isn't just about being alone but stems from the inner numbness and emotional void caused by excessive external pursuit in the fast - paced life.

When young people strive in society under great work pressure, they often neglect their emotions. When I worked in big tech companies before, I often ignored my own feelings and hid many emotional needs just to meet external expectations and achieve worldly success.

But at a certain critical point, I suddenly realized that this state was wrong. Many friends advised me to "relax". By then, I had lost the ability to feel love and felt extremely empty and lonely.

At this time, I found that the best cure for such moments was to raise a pet. Through the process of caring and giving, I could gain a sense of satisfaction. Ropet is actually designed to provide better solutions for these "weak moments".

"Intelligent Emergence": Your previous jobs were all about functional products. Is it difficult to create a very emotion - oriented product now?

He Jiabin: Actually, I've always been working on emotion - oriented products. This opportunity just allows me to fully unleash my creativity.

I graduated from the Beijing Institute of Fashion Technology. I once designed some interesting products, such as a shared Wi - Fi backpack. The idea behind that product was to bring people closer. Many of the cross - border product designs I did in school were aimed at solving social and companionship problems.

When I first co - founded the company to make the Luka picture - book robot, I was also creating a product with emotional value, although it was mainly for children. I'm quite comfortable and happy when creating fun and cute things.

However, the product attribute of Luka, where the purchaser and the user are different, made me feel a bit conflicted during the process. To increase sales, we had to convince parents of its tool value, and to enhance user stickiness, we had to make the product fun for children. Good design can somewhat balance this issue, but this problem more or less exists in children's early - education hardware.

It wasn't until this startup that I no longer had to compromise on other business considerations. I could completely focus on creating a product that provides emotional value and that I like, allowing me to more freely create new things.

△Image source: Company authorization

"Intelligent Emergence": Your starting point is quite different from that of entrepreneurs who come from the AI large - model technology field to create similar products.

He Jiabin: Yes. Founders with a technical background may be more concerned about what new things AI can bring, whether large models can quickly transform an industry, and whether commercialization can grow rapidly.

But I think the most important thing is that the product needs to have real - world scenarios and be liked by users. Currently, AI only helps us solve some "minor problems", while good design can solve major problems. If a product has a certain user base, real - world scenarios, good stickiness, and a large amount of user "in - the - moment" data, AI will help solve industry - wide challenges.

Good aesthetic design can quickly establish a connection between the product and users. Ropet has a design that users (especially female users) will like at first sight. I believe I have a unique perspective on design, user experience, and user aesthetics in this field.

The founder's subconscious will be internalized into the company's management decisions and product strategies. The differences between a technical background and a design background will lead to completely different product development paths.

The non - talking AI pet: Nearly ten thousand units sold in half a year

"Intelligent Emergence": Are you worried that it will become a niche market and that educating users will take a long time?

He Jiabin: We're not particularly concerned about educating users. The need to raise pets is a common human trait. I believe that in the future, users of all ages and different cultural backgrounds may solve their emotional problems by raising virtual pets, and every family will have their own AI pet robot.

Currently, pet robots have some problems: a lack of good design and a lack of psychological research and AI integration technology to create machine creatures with a sense of growth and generalized behaviors. If we can use existing technological means and good design capabilities to rebuild a more easily - raised pet category, market demand won't be an issue.

"Intelligent Emergence": So Ropet's product logic is similar to that of a pet.

He Jiabin: Yes. The core of Ropet's "content" is to deliver a cute robot to users, which can stimulate their motivation to take care of it. This process of taking care is gamified, gradually creating a sense of exclusivity and an intimate relationship.

The core logic of raising a pet is the same. When I raise an American Shorthair cat, I stroke it every day and feed it cat treats. Its fur becomes shinier, and it becomes more attached to me, thus establishing an intimate relationship.

△Image source: Company authorization

"Intelligent Emergence": What are the characteristics of the companionship needs of women?

He Jiabin: We conducted more than a dozen surveys on our target user group of urban women aged 20 - 40 in different regions around the world. We had questionnaires with over 1000 samples and in - depth interviews with dozens of users. Most importantly, we also have more than 1000 real users continuously using the product, providing real - time data feedback.

We discovered some non - consensus findings, such as conversation. For women, the scenario of chatting with an AI toy or AI pet for companionship doesn't hold true. Most of the time, they're reluctant to talk to robots, except in two situations:

One situation is to explore the boundaries of AI through continuous questioning; the other is for the tool - based function, such as when users search for knowledge nearby and hope to get quick results. These two scenarios aren't necessarily related to companionship and solving emotional problems. They're more about novelty and practicality.

Our conclusion is that companionship doesn't necessarily mean conversation. If we want to create a cute and unique hardware product, it definitely doesn't rely on the large - model's conversation ability. The difficult part and the competitive edge lie in how to make it more like a real living being with biological perception and stress responses.

"Intelligent Emergence": How does Ropet achieve perception and stress responses?

He Jiabin: The "pet brain" decision - making model we're developing is based on a Transformer - architecture end - side multi - modal model (including vision and voice) training. Its main function is to process user behavior information and give accurate emotional responses.

This decision - making model is Ropet's exclusive and core competitive edge. If a startup can only rely on the large - model conversation ability provided by big tech companies, its competitive edge will be weak, and there will be a high degree of product homogeneity.

"Intelligent Emergence": Is there a learning curve for users to start using the Ropet product? Can they understand its value right away?

He Jiabin: There is a certain learning curve. Some users have a stereotyped impression of "companion robots" and think that companionship must mean being able to talk about anything.

Second, since the product is positioned as a robot, some users assume that it can move. So these users may not be the early adopters of Ropet's product form.

However, some Chinese users were very interested after seeing the product promotion. They asked their overseas friends to place orders on the crowdfunding website (it hasn't been launched in China yet). After waiting for four months with great difficulty, they finally received the product. They're very tolerant of the product's imperfections during the process, and we hope to listen to their suggestions for product iteration and further improve the product.

"Intelligent Emergence": How do you create a good user experience for your target users?

He Jiabin: First, we deliver the product and collect users' feedback on functions, their expectations, and interesting moments of interaction with Ropet. Then we encourage users to share their interaction experiences in the community to help other users better explore the growth - oriented gameplay.

At the same time, users' feedback also helps us to iterate the product more clearly. Our ongoing "Brain Co - creation Project" is a very interesting community activity where users can co - create.

△Ropet community sharing   Image source: Company authorization

"Intelligent Emergence": Have you received any unexpected feedback?

He Jiabin: Some users have reported that after raising Ropet for seven days, it developed its own personality and recognized its owner. When the device was accidentally damaged and the data was cleared, they felt like they had lost something and hoped to restore the data.

Some users actually play house with R