Recently, the Ministry of Industry and Information Technology issued a document proposing to "encourage service providers to build Forward Deployed Engineer (FDE) teams." Data from recruitment websites shows that demand for FDEs has been strong since 2026, with job postings in August alone exceeding the total for the first half of the year. Industry insiders say the core work of an FDE is not only to bridge the "last mile" of AI deployment, but also to demonstrate results with quantifiable outcomes. Given the high entry barrier—it is hard to master quickly and requires a mix of technical, business, consulting, and project management skills—FDEs are known as the "special forces" of the workplace. With supply falling short of demand, research institutions, universities, and open-source communities have already begun FDE research and training, and relevant standards and cultivation systems are being developed and refined.
What Is an FDE
The Ministry of Industry and Information Technology recently issued a notice on launching a special initiative to cultivate artificial intelligence application service providers, which proposes to "encourage service providers to build Forward Deployed Engineer (FDE) teams that take root at customer sites to ensure scenario implementation."
On what an FDE is, several industry insiders told reporters that FDEs are known as the "special forces" of the workplace; they must combine engineering development, business understanding, and client communication skills, and be able to turn vague requirements into executable solutions.
The AI Research Institute of the China Academy of Information and Communications Technology believes that the Ministry's notice marks FDE's transition from industry exploration to policy deployment, making it an important lever for enhancing AI application service capabilities.
Wang Zhilan, an FDE expert in retail, gave reporters an example: she placed an AI agent in a retail company's work group chat, letting AI observe and summarize the company's chat records, and it uncovered many problems that even the company's owner had not noticed. For instance, the company has many products, and every time packaging is changed, content such as nutrition facts tables, compliance descriptions, and channel requirements must be reviewed. Previously there was no unified standard for packaging review, so many similar errors kept recurring, and staff could only communicate back and forth in the group, often waiting three to five days.
To address this, Wang Zhilan first broke down the workflow step by step, then developed corresponding AI agents so that humans and AI each handled the work they were best at. Subsequent verification showed that 80% of low-level errors were intercepted by AI, communication time was cut by more than 50%, and the queue for packaging review essentially disappeared.
"This is just a small example. An FDE's job is to clear all kinds of bottlenecks and deliver quantifiable results in cost reduction and efficiency gains," Wang Zhilan told reporters. An FDE does not deliver a single project or piece of software, but rather a redefined "human + AI" collaborative organization.
"In the future, every company will need the FDE role," said Chen Yu, an FDE expert and co-founder of Yuanzhen Technology. "They can be provided by external service teams, or by internal operations staff with development capabilities who transition into the role."
The Smarter AI Gets, the Scarcer FDEs Become
Today, FDEs are seen as a key role in helping AI applications succeed commercially. Everywhere you look, from AI companies to all industries, everyone is embracing FDEs. Recently, the "2026 Yangtze River Delta FDE Joint Action—FDE Enterprise AI Deployment Class" hosted by the AI open-source community Datawhale attracted hundreds of entrepreneurs to sign up. Follow-up visits within a week after the event showed that over 80% of connections between FDE experts and target companies had been completed.
FDE hiring demand has entered the fast lane. Data from recruitment websites shows that since 2026, new FDE-related job postings have climbed month by month, with August's postings alone roughly equal to the total for the first half of 2026. June saw a month-on-month increase of 237.21%, and July and August continued to grow on that basis. Over the past year, the average annual salary for new FDE-related positions on Liepin was 408,000 yuan, with a median of 366,000 yuan—both significantly above the all-industry average. By distribution of FDE-related emerging roles, AI, internet, and IT industries account for 72.36%, the most concentrated areas of demand. In addition, FDE demand from healthcare, transportation and logistics, finance, automotive, and other industries is growing, and FDE job distribution now extends beyond the tech industry.
Looking ahead, Guo Wenwei, founder of industrial large model company Lingnum Technology, believes FDE demand will grow rapidly.
How FDEs Are Made
Despite strong demand and handsome pay, the industry generally agrees that FDEs cannot be trained quickly.
Chen Yu told reporters that FDEs face high comprehensive capability requirements and are difficult to cultivate. They must be able to clearly calculate return on investment for clients, and also possess product and operations thinking; they must not only continuously track product usage, but also iterate and optimize solutions based on user feedback and business fit, avoiding problems like system complexity and redundant construction. The most important quality of an FDE is still the ability to deliver results—helping clients improve core business metrics such as customer acquisition scale and conversion rates, completing the upgrade from digitalization to intelligence.
What backgrounds make it easier to become an FDE? After recruiting and screening nearly 50 FDE experts, Si Yuxin, head of the Datawhale ecosystem, analyzed their professional profiles and found that their résumés span major internet companies, research teams, and industry frontlines; most are technical leads, product managers, and engineers, distributed across manufacturing, e-commerce, software and internet, finance, and professional services industries. They generally have enterprise-side business backgrounds, and some come from delivery teams at startups or established SaaS companies.
Now, FDE talent cultivation has been put on the agenda. The AI Research Institute of the China Academy of Information and Communications Technology plans to advance a series of FDE initiatives in four areas—theoretical consensus, standards systems, evaluation and certification, and industry ecosystem—gradually forming a work system that runs from concept research to standards implementation, from capability assessment to talent cultivation, and from case accumulation to ecosystem promotion. Recently, Shanghai Jiao Tong University's Antai College of Economics and Management, together with Datawhale and other industry ecosystem partners, also jointly launched the "FDE Talent Specialized Cultivation Program."
On the FDE talent growth path, Ba Ran, Vice President of Tongdao Liepin Group, believes that AI deployment talent usually needs to accumulate experience through real projects and cannot be trained in batches in a short time. Rapidly released demand and the time required for talent cultivation give FDEs with mature delivery experience greater bargaining power. From a talent supply-and-demand perspective, AI talent opportunities are extending to people with industry experience and project delivery capabilities. Talent in software development, solutions architecture, technical consulting, and other fields may all enter relevant roles by supplementing their AI engineering skills. Data, systems, and business rules differ across industries, and talent who can understand these differences and turn them into executable solutions are more likely to realize the value of their hybrid capabilities.
"FDEs are hard to train quickly, but their development can be accelerated," Guo Wenwei suggested. Engineering and AI skills can be acquired through systematic training, but industry insight must be soaked up and accumulated in real projects; the proven path is "expert mentoring + real-scenario practice + retrospective consolidation."
(Source: Economic Information Daily)
Solemn statement: East Money publishes this content to disseminate more information. It does not represent this site's position and does not constitute investment advice. Anyone acting on it bears the risks themselves.
Scan the QR code to follow
East Money official website on WeChat
Microsoft's earnings report shines: Azure surges 48% year-on-year
