Summary of Key Points
FDEs (Frontline Deployment Engineers) are a new profession that has emerged in the AI era, responsible for implementing AI solutions on-site at companies. However, this role is currently in an "embarrassing" phase: their identity is unclear (customers don't know exactly how to refer to them), delivery standards are chaotic (some provide demos while others sell complete systems), frontline employees are resistant (fearing they will be replaced by AI), and industry pricing is absurd (some charge as much as 2000 yuan for a project that costs only 190,000 yuan). In the long run, FDEs are similar to the "system analysts" from the computer普及 era of the last century; they will play a crucial role in making AI a true source of productivity for businesses, and the current chaos is just temporary.
1. What are FDEs, and why don't even customers know how to address them?
FDE stands for "Frontline Deployment Engineers," which essentially means "people who bring AI technology to companies to help implement solutions." But this role is so new that even customers are unsure of its exact role:
- Mai Mai, an FDE from an AI company, had to pretend to be a "product manager" when visiting a client site; Xie Yao, a freelance FDE, was asked which department he belonged to during overtime and could only say he was an "outsourced worker."
- Previously, these roles were called PEs (Project Engineers). It was so difficult to recruit for FDEs that the name was changed recently, although the tasks remain the same.
The fundamental reason: FDEs do everything—from understanding the business to building systems, coordinating requirements, to maintaining them—there is no fixed label, so customers have to assign them temporary roles based on their needs.
2. Two types of FDEs: the "regular army" with their own products vs. the "guerrilla force" that relies on connections
There are two types of FDEs, with completely different approaches:
1. Product-driven (regular army):
- Origin: Large companies (such as those founded by OpenAI or Anthropic) or companies with their own AI products.
- Approach: They bring their own models/platforms to clients, customize solutions, and use customer data to improve their products (e.g., identifying weaknesses in certain industries).
- Career path: Beginners act as outsourced workers, but over time, they gain experience and can provide advice to clients (e.g., suggesting "This method isn't effective; let's test it with data.")
2. Project-driven (guerrilla force):
- Origin: Small teams or individual FDEs who obtain work through referrals or industry connections.
- Approach: They sell their expertise and time, such as by developing systems or providing consulting services. The downside is that they need to learn new things for each industry, which leads to increased after-sales work and a risk of becoming mere outsourced workers.
Key difference: The regular army has the support of their products, allowing them to build experience; the guerrilla force is more flexible but has limited potential and is easily influenced by client demands.
3. Is the client environment like a "drama?" Why are frontline employees resistant to FDEs?
The biggest issue for FDEs is not technology, but people:
- Employee resistance: Frontline employees see FDEs as outsiders hired by management to "optimize processes," often with hostility, and they may deliberately slow down projects (e.g., a development manager preventing the use of advanced AI models).
Solutions: Sun Wuyuan, a freelance FDE, adopted a strategy: first explain to the employees that he aims to improve efficiency, not just replace them; then clarify the role of AI to management to avoid extra burdens; finally, find a few willing employees to test the benefits of AI (e.g., showing reduced repetitive work) to gradually overcome resistance.
Core skill: Being observant and understanding the internal dynamics of the company, knowing what management really needs, and managing conflicts between employees and management—it's like navigating a "drama."
4. How chaotic is the industry?
The FDE industry is in a state of "wild growth":
- Absurd pricing: The same project can be priced differently significantly (e.g., 190,000 yuan for 3 months vs. 2000 yuan for 10 days). The lower price may only cover the cost of the model subscription (1500 yuan), without considering additional modifications and maintenance.
- Non-uniform delivery standards: Some provide just a demo, while others deliver a fully functional system; some offer training, others require long-term on-site presence.
- Chaos in pricing models: Fees can be based on time, projects, or results, with different levels of risk, but customers only focus on "speed and price."
Consequence: This leads to even less trust from frontline employees: "Can you really help me?"
5. Is FDE a trend or a fleeting phenomenon? Look at history
FDEs didn't appear out of nowhere; similar roles have existed before:
- In the 1950s and 1960s, when large computers entered companies, "system analysts" emerged—people who understood both the business and technology, translating vague requirements into actionable systems for programmers. This role was initially unclear and overlapped with those of programmers and consultants, but it became essential for businesses.
- FDEs are currently in the same phase: helping companies implement vague AI requirements. Although chaotic, they will eventually become a critical role for AI productivity.
Conclusion: Those who enter the FDE field now can reap the initial benefits, but they must endure the chaos. In the long run, this role will become as indispensable as system analysts did for businesses.
Final Summary
FDEs are the "bridge" for implementing AI in companies. The current chaos is a "growth pain." Once companies learn to clearly define their needs, FDEs can establish clear roles, and the industry will set standards. For those considering entering this field, be prepared to combine technical expertise with the ability to navigate complex corporate dynamics.