The new structure

哈啰 has completed an internal reorganization of its Robotaxi business unit, as confirmed to the agency 蓝鲸汽车 according to what IT之家 reports. The teams previously part of 智能车 — model development, system solutions and evaluation, as well as the AI Infra data platform — have been merged into a new department called "模型开发与数据闭环部", that is, model development and data closed loop. It is directly headed by CEO 杨磊, with an internal structure divided into one department for model development and one for the data closed loop.

The choice to place responsibility at the CEO level signals that the function is considered central and not delegated to an intermediate level. At the same time, activities relating to AI-native vehicles, the second generation of vehicles and engineering delivery for production move under CTO 于乾坤. Otherwise, the source indicates that the architecture of the other departments of 智能车 and their reporting lines remain unchanged.

Talent and funding

The reorganization is accompanied by a personnel move: 哈啰 has confirmed that 李一贤, previously at the embodied robotics company 无界动力, has joined Robotaxi. The news had circulated as a rumor about the addition of notable algorithmic expertise; the company confirmed its accuracy without providing further details on the role held.

On the financial side, the material recalls the project's already known milestones: in June 2025 哈啰 had announced its entry into the Robotaxi sector, with the establishment of 造父智能科技有限公司, an initiative promoted by 哈啰 together with 蚂蚁集团 and 宁德时代 through their respective investment vehicles. In September 2025 the company had announced a strategic investment by 阿里巴巴 in the Robotaxi business. Also in September, Robotaxi completed a round of about $100 million, led by 上海国投先导基金 with the participation of 闵金投 and 创世伙伴创投, at a post-money valuation close to $3 billion. According to the source, the funds will be destined mainly for research and development, with a focus on cluster computing platforms of tens of thousands of cards, end-to-end iteration of AI models and the construction of a high-quality data supply chain.

Why it matters and what remains to be clarified

The merging under a single department of model development, system evaluation and data platform makes explicit an operating model in which data quality and model training are treated as a single cycle. It is a choice consistent with the spending priorities indicated by the company — compute, end-to-end models and data sourcing — and with the confirmation of the arrival of an algorithms specialist. The separation between this core, led by the CEO, and the vehicle and production activities, entrusted to the CTO, outlines a division between those who develop the system's intelligence and those who bring the vehicle into production.

However, some elements not covered by the material remain to be verified: it does not state how many people are involved in the new department or what 李一贤's exact responsibilities are; timelines, measurable objectives or milestones of the Robotaxi program are not specified. The source also does not report comments from the managers involved or details on the relationship between the new structure and 造父智能's corporate framework. The information reported here comes from the company confirmation cited by IT之家 and should be read as such.

Source and attribution: this original summary was prepared by AICronaca based on the report by IT之家 ↗Statements are attributed to the source; any information limitations: The material is based on a company confirmation reported by IT之家 via 蓝鲸汽车 and does not contain data on headcount, specific roles, timelines or program objectives; the financial references are those already disclosed previously and are not updated for this reorganization.