The ability to quickly assemble a highly capable technical team has become a critical factor determining whether embodied AI companies can achieve commercial success first. Wang Xin (a pseudonym), an algorithm engineer specializing in VLA and world models, recently received job offers from three different headhunters for roles in the embodied AI sector, making him a sought-after technical talent. Attracted by the promised salary and career prospects, Wang Xin chose to leave a major internet company to join a unicorn startup. Due to a mismatch in the supply and demand market, the salaries for top-tier talent and technical backbone staff in embodied intelligence have been driven even higher. "Candidates who are skilled in algorithms, have some experience with physical robots, and have worked on high-value projects can likely command an annual salary of around one million yuan," Wang Xin added.
Since the beginning of this year, as financing in the embodied AI track continues to heat up, companies ranging from industry giants to startups have joined the fierce competition for talent. The ability to rapidly build a strong technical team is now a key variable determining which embodied AI firms will be the first to commercialize their technology successfully.
Starting last year, the embodied intelligence sector has noticeably warmed up, with an increase in financing events and the number of related companies. This year, state-backed entities, industrial capital, and top-tier venture capital firms have all entered the field, betting on embodied AI, leading to the emergence of a cluster of unicorn companies. This boom has transferred to the recruitment market, driving up demand for related positions. According to a report from Liepin Big Data Research Institute, the robotics field is in a stage of rapid expansion coupled with an upgrading demand for talent. In the past year, new job postings in this sector grew by 75.26% year-over-year. Data from Zhaopin shows that in the first quarter of this year, job postings in the robotics industry grew by 38%, with positions for industrial robotics engineers, robotics algorithm engineers, and robotics commissioning engineers growing by 38%, 37%, and 60% respectively.
"The embodied intelligence industry experiences fast technological iteration; currently, VLA and world models are quite hot," Wang Xin stated, noting that there are not many talents in the market familiar with these technologies and directions. A company referred to as Xinghai Tu also indicated that as their business scale continues to expand and their product lines diversify, algorithm engineers and other core R&D technical personnel are in short supply. Competition for high-quality technical talent within the industry has always been relatively intense. Both leading enterprises and startups are increasing their technical investments, intensifying the competition for talent. This has pushed up overall recruitment costs and salary levels for core technical talent, with market salaries showing a clear upward trend influenced by supply and demand.
The report shows that the average annual recruitment salary in the robotics field over the past year reached 328,000 yuan, which is generally at a high level. A review of job postings from several leading companies and startups in the industry reveals that technical positions often require a master's degree. The starting monthly salary for an embodied intelligence algorithm engineer is mostly around 30,000 yuan; for an expert-level position, the starting monthly salary is 50,000 yuan, while a world model algorithm engineer can command a starting monthly salary of 60,000 yuan. Other popular positions include motion control algorithm engineers and embedded software engineers.
Wang Xin's most recent position is as an algorithm engineer. "At the time, one headhunter recruiting for a startup offered a salary that was 150% higher than my previous role. The other two offers ranged from 50% to 100% increases," he said. After weighing his options, he ultimately did not accept the offer with the highest salary. "I placed more importance on whether I could access more computing resources and if talent support was sufficient; salary wasn't the most important factor at that time," Wang Xin revealed. He added that to compete for talent, some startups are currently offering higher salary packages than leading companies. "The embodied intelligence industry is still in its early stages; candidates with just 1-3 years of experience in algorithms are considered highly desirable."
Zhang Hao (a pseudonym), a headhunter familiar with the hiring needs of robotics companies, supplemented that generally, senior engineers can achieve an annual salary in the millions, with heads of business lines earning even more, including signing bonuses, relocation fees, and stock options. Recruitment information from multiple companies further reflects this strong desire for core technical talent. For example, on April 2nd, a company announced a global recruitment drive for a "Chief Scientist of Embodied Intelligence," with an annual salary starting at 15 million yuan and potentially reaching up to 124 million yuan. Last year, a team from another major tech firm initiated a high-specification recruitment seeking a "Senior Expert in Operational Algorithms (Embodied Intelligence)," explicitly focused on the development of "humanoid embodied robots," with a monthly salary ranging from 95,000 to 120,000 yuan.
Despite the premium on salaries in the embodied intelligence industry, many job seekers are also looking for opportunities to transition into the field. According to interviewees, job seekers entering the embodied AI industry primarily come from three backgrounds: major internet companies, smart driving/autonomous vehicles, and smart manufacturing. In his spare time, Wang Xin shares entry-level experiences on the social platform Xiaohongshu under the pseudonym "Silicon-based Stroll" and has communicated with numerous job seekers. He explained that companies, whether hiring new graduates or experienced professionals, place greater emphasis on project experience during recruitment. "Companies are struggling to find people; even with high salaries, they can't find versatile talents who understand both algorithms and physical hardware. Job seekers are also frustrated because a lack of project experience makes it difficult to enter the industry," Wang Xin noted, pointing to a current mismatch in the embodied AI recruitment market.
It has been observed that many companies prioritize candidates who understand technology application and implementation during hiring, where publishing papers at top-tier conferences and having hands-on debugging experience with physical robots are significant advantages. Feedback from Xinghai Tu indicated that they find mature talent possessing both professional technical skills and practical industry experience to be relatively scarce during their recruitment process, requiring repeated screening to find suitable candidates.
Wang Xin specifically mentioned that the increased demand for motion control engineers at some startups might be aimed at quickly developing demos to present to VC firms or investors. Zhang Hao also recounted that some startups in the industry seek talent who can "get the job done" and tend to prefer recruiting experienced professionals. He stated that based on a company's development stage, its urgent hiring needs can change. Once a product becomes more mature, some companies might prioritize sales and business development roles.
Furthermore, the aforementioned report notes that the talent structure in the robotics field is evolving from a "single-technology orientation" towards "compound capabilities combining technology, engineering, and business," with interdisciplinary integration becoming a core trend for industry development. To alleviate the shortage of technical talent, several companies have announced partnerships with universities to cultivate talent. This could be a way to lock in talent early and build a talent moat. Additionally, universities such as Beihang University, Beijing Institute of Technology, and Shanghai Jiao Tong University have already established undergraduate programs in "Embodied Intelligence."
"Embodied intelligence has high requirements for computing power, and many universities lack the resources to cultivate such talent," Wang Xin judged. He believes that university-corporate cooperation and academic training will take some time to yield results, and the talent shortage in the embodied intelligence industry is likely to persist in the short term.
Comments