According to the Zhitong Finance APP, TONGDAO LIEPIN (06100) released its interim results for the six months ending June 30, 2025, with group revenue of RMB 937 million; the net profit attributable to equity shareholders was RMB 80.3 million, a year-on-year increase of 77.13%; basic earnings per share were 16.82 cents.

In the first half of 2025, based on deep insights into the recruitment market and extensive research on cutting-edge technologies, the company continued to promote the application of products around "AI + Recruitment," establishing and continuously enriching its AI recruitment product matrix. At the same time, the company's sales strategy focused on the fully transformed "LIEPIN AI Enterprise Edition" product. By concentrating on new products and a flat organizational structure, the company significantly enhanced its market competitiveness, achieving good growth in newly signed enterprise clients. During the reporting period, the group restructured the value system of its enterprise client resources, utilizing self-developed AI tools for lead management, resource classification, and formulating targeted sales strategies, which improved sales efficiency while rapidly strengthening enterprise client penetration. By the end of the first half of 2025, the total number of registered enterprise users reached 1.43 million, a year-on-year increase of 4.8%. The number of paying enterprise clients was 62,900, with a year-on-year decline narrowing to 0.9%.

To continuously enhance customer service capabilities, LIEPIN insists on product innovation and seizing opportunities in the AI era. In the first half of this year, the "LIEPIN Enterprise Edition" implemented AI account upgrades, offering new features such as "AI Resume Quick Read" and "AI Intelligent Invitation." Together with the core product launched in the fourth quarter of 2024—"Intent Communication," a relatively complete AI recruitment product matrix was formed, covering various needs of recruiters from efficiency improvement to result delivery. In the first half of 2025, AI recruitment products covered over 70% of LIEPIN's paying clients, a significant increase from the beginning of the year. Benefiting from the growth in AI matching capabilities, the recommendation rate for potential candidates increased significantly to 64% within two hours, and the monthly repeat initiation rate of the product continued to grow and stabilize above 60%, fully reflecting the high recognition and deepening usage habits of enterprise clients for the company's AI recruitment products.

On the basis of the comprehensive upgrade of the "LIEPIN AI Enterprise Edition," AI has also brought more diversified business growth and empowerment to LIEPIN. The LIEPIN AI Intelligent Interview System, Doris, has received high market attention since its launch in 2024. Based on LIEPIN's deep cultivation in the recruitment field and continuous technological iteration, Doris possesses unique technical capabilities in the intelligent interview field, covering interview needs across different job categories, from white-collar to blue-collar and gray-collar recruitment. The consistency of AI system evaluation results with those of experienced interview experts has remained stable at over 95%, providing a solid guarantee for enhancing user experience. In the first half of 2025, this business focused on expanding across multiple scenarios and breaking through industry benchmark clients. The company is currently concentrating on establishing brand reputation and industry service capabilities through leading benchmark clients, laying the foundation for rapid replication in the next phase. On the other hand, based on the capabilities of the AI intelligent interview system, the company can not only meet the needs of enterprises for bulk interviews but is also applicable to internal talent assessments, business evaluations, promotion assessments, and other scenarios. The diversified usage scenarios are expected to open further market space for this product AI technology has not only empowered core products but also brought significant improvements to the headhunting collaboration network platform (DuoLie RCN) business and the closed-loop delivery business under Liepin. DuoLie RCN is committed to enhancing the delivery capabilities of headhunting companies within the network and strengthening their competitiveness in the market by providing a business SaaS system and a collaboration platform among headhunting firms. As of June 30, 2025, the number of positions in operation within the RCN business has reached 159,000, achieving continuous and stable growth. Currently, the embedded AI functions within RCN have helped headhunters with core tasks such as talent search, matching assessment, and intention communication. On this basis, in the third quarter of 2025, the company will launch the order assistant Xiao Yi (Agent), which will facilitate a smoother and more convenient order delivery process for headhunting firms and further deepen the collaborative stickiness between Liepin and headhunting companies.

The rapid iteration and upgrade of products are attributed to Liepin Group's emphasis on technological research and development. In the field of AI algorithms, the group extensively accesses and deeply utilizes various mainstream large language models and has deeply trained its self-developed vertical model— "Tongdao Huicai," thereby constructing a multi-level architecture that includes core model layer, technology platform layer, and business application layer, forming a complete closed loop from technological barriers to business implementation. In the first half of this year, the company fine-tuned its self-developed model through distilling large language model data and, based on this, utilized closed-loop data at the level of hundreds of millions of tokens for industry vertical data reinforcement learning, significantly enhancing the reasoning ability of its self-developed model. At the technology platform layer, based on its understanding of the recruitment process and the accumulation of matching logic, the company injected industry insights into the large model through retrieval-augmented generation (RAG), continuously improving the accuracy of job matching. Since the beginning of this year, through continuous experience accumulation and model optimization, the number of negative feedback from key industry enterprise users on AI products has decreased by 50% year-on-year. Finally, based on model capabilities, the company further combined user needs at the business application layer, gradually launching multiple AI application products covering enterprise, headhunter, and candidate (BHC) users, creating user value and providing a complete AI solution for the entire recruitment process