36Kr Exclusive: Focusing on "AI + Tourism", "ShiLv Technology" Secures 100 Million Yuan in Series A Financing

36Kr has learned that "Vision Travel Technology" recently announced the completion of a Series A financing round of 100 million yuan. The investors include Shenzhen Daily One Fund, the Cultural Industry Investment Fund of the Guangzhou City Investment Group in the Guangdong-Hong Kong-Macao Greater Bay Area, Gaorong Capital, Youshan Capital, and Zhonghuyakang, the investment platform of Yaskawa Electric (China) Co., Ltd., with a total scale of over 100 million yuan. China Renaissance served as the exclusive financial advisor for this round of financing. This round of funds will be used to accelerate the research and development and upgrade of the VtripGPT tourism large model, and continue to promote its market-oriented application in high-quality tourist destinations at home and abroad.

("Vision Travel Technology", hereinafter referred to as "Vision Travel") was established in November 2021. It hopes to build an "AI-driven new-generation tourism platform" based on a large model, focusing on various pain points of consumers during travel, such as the cumbersome process of planning a trip, the complexity of prices and strategies, the popularity of itineraries, the inability to have in-depth experiences, and the lack of high-quality sharing materials. It combines AI to optimize the travel experience. Currently, its products for C-end users, such as "Lepai.com" and "Golden Panda Travel", have been able to realize functions such as AI itinerary design, dynamic packaging and booking, and intelligent travel photography.

Highlight Display of the Application of Vision Travel Technology's VtripGPT Tourism Large Model

Looking back at the process since its establishment, Vision Travel spent nearly a year in 2022 building the basic capabilities of the AI bottom layer and expanding the core destinations in the initial stage. Subsequently, along with the AI wave set off by ChatGPT, it released the first large model in the domestic tourism field, VtripGPT1.0, in May 2023. The AI user products were also released simultaneously and landed in Hong Kong. From August to September of the same year, the AI user products in destinations such as Singapore, Ganzi, Manba, Banna, and Chengdu also landed one after another.

Behind Vision Travel is a star team with experience in both the tourism industry and AI technology. The founder, Li Shaohua, was once the key person who built "Fliggy". The CTO, Hui Pengrui, is a doctor of mathematics and was also a former senior technical expert in artificial intelligence at Alibaba.

The Main Creative Team of Vision Travel's Large Model (From left to right are CTO Hui Pengrui, CEO Li Shaohua, Vice President of Technical Products Liu Runchun, and Head of AI Large Model Research and Development Chen Xia)

In the early stage of its establishment, Vision Travel received an angel round investment of 100 million yuan from investment institutions such as Gaorong and K2VC. When talking about the changes that AI has brought to the tourism industry, Hui Pengrui concluded, "What AI does is to deeply connect users and destinations, providing each user with an exclusive customized travel experience, and at the same time forming a scale fulfillment advantage at the destination end, thereby truly solving industrial problems."

From the perspective of the establishment time, Vision Travel was born in the window period of adjustment and reconstruction of the tourism industry. On the one hand, the world as a whole has entered the post-epidemic era, and the tourism industry has turned from cold to hot, with obvious consumer confidence and growth trends. On the other hand, the online trend of tourism is becoming more and more irreversible. In 2019, the online proportion of the industry was only 27.3%, and as of October 2024, this figure has exceeded 60%. Hui Pengrui said, "Under the online wave, there are many possibilities for innovation."

The focus of Vision Travel is clearly aimed at one direction - "Online transactions for destinations". Generally speaking, in the tourism business, one end is the consumer, and the other end is the destination resources. "Through AI technology, we integrate the fragmented product end-to-end and provide it to users in the form of a dynamic customized package."

In the previous report of 36Kr on Vision Travel, the approximate usage experience of the user end was mentioned: Put forward requirements to the AI butler - Automatically plan the itinerary - Generate the product combination required for the itinerary - Place an order and fulfill it on the ground. During this process, the itinerary can be interactively modified and detailed at any time. In other words, it is neither to choose from the fixed product packages pre-combined by the platform merchants, nor do you need to search and book one by one. Instead, it directly connects the various experience products of the destination to the users and adjusts the combination at any time.

Since its establishment, the Vision Travel team has been considering and promoting the business around three core points:

First, the starting point and goal. Hui Pengrui said that what Vision Travel wants to do is very certain, that is, to build an end-to-end collaboration between consumers and destinations. Combining AI technology, the rich supply of the cultural and tourism industry is directly and conveniently pushed to consumers.

Second, the core basic capabilities. Currently, Vision Travel's business is built around the tourism large model VtripGPT it has developed. The overall architecture of the entire model is roughly that first, the most advanced general large model is applied at the bottom layer to train the exclusive AI base capabilities. On this basis, a large amount of tourism domain data, expert knowledge, and real-time user interaction feedback data are introduced, and iterations are formed through a reinforcement learning mechanism.

Third, based on the goals and capabilities, explore the possibilities of AI applications. The core question is, "With a vertical large model and targeting the industrial scenario, what role does Vision Travel play in it?" Based on this, Vision Travel directly connects and technically processes a series of data such as the product library, itinerary library, POI library, and content library with the large model, and provides product services for the final user scenarios.

The Overall Model of Vision Travel

Up to now, several directly-operated destinations that Vision Travel focused on in the early stage have achieved initial results - including Sichuan, Hong Kong, and Singapore. The monthly fulfillment scale of some core destinations exceeds 10 million yuan, and the business model has basically been verified. Since the second half of 2023, Vision Travel has begun to establish a digital operation cooperation system for destination travel agencies, cooperating with partners who are deeply rooted in one place to expand destinations, and the corresponding expansion speed is also getting faster and faster.

According to Hui Pengrui, Vision Travel's revenue in 2023 is close to 200 million yuan, and there are about 1 million users who have interacted or fulfilled in the products. This year, it is expected to reach 3 million person-times.

It is worth mentioning that currently, about 85% of Vision Travel's revenue comes from commission income, that is, the income brought by users' bookings on the platform. In addition, there is also a portion of revenue from the B-end, cooperating with the tourism development departments or scenic spots at the destination to provide services such as market promotion, exclusive agency operation, and AI product development.

In October 2023, Vision Travel won the Gold Award in the 2023 National Finals of the Huawei Ascend AI Innovation Competition; in March this year, their VtripGPT passed the first tourism large model filing in China. Hui Pengrui said that Vision Travel takes "Building a Bridge for Hundreds of Millions of Chinese People to Go to the World" as its mission - "Wherever there is a demand in the tourism scenarios for Chinese people, there is our space and value."