Unlock Services

After logging in, more features will be unlocked. Come enjoy personalized services!

2026.09.29 | 3037 views

0 CommonWealth Magazine

2026.09.29

3037 views

Why has the Forward Deployed Engineer (FDE) become a hot new job opening? Companies from Dcard and Hotai Motor to Pegatron, Foxconn, and Wistron are all hiring. Generative AI is accelerating its way into enterprise workflows, and new positions are emerging along with it—from FDEs who help companies deploy AI agents to Chief AI Officers in charge of governance. How is the corporate org chart being rewritten?

Text / Kang Chiung-chih Images / Hsieh Kuan Republished with permission from CommonWealth Magazine

Table of Contents (click to jump)

Lee Kun-che, a senior PM with a finance degree who can't write code, never imagined he would one day do work so close to engineering—helping companies build AI agents.

At the end of May this year, amid the generative AI wave, Taiwan's largest digital-native community, Dcard, announced the establishment of a new business unit, "GNTC."

The goal is to productize its hands-on experience driving AI transformation internally, helping companies adopt AI agents that connect to existing internal systems and unify data permissions and workflow management.

He had originally been responsible for exploring the company's "second growth curve" at Dcard, and as a result has now shifted his focus entirely to AI.

Lee's work starts by collecting each department's requests for "what can AI do for me." Some people spend too long making weekly reports, some want to find information in databases, and some want to quickly create client proposals. He and his team break down the workflows and help colleagues turn each request into an agent.

Today, the company has accumulated nearly 700 agents internally. This isn't just about employees having handy AI tools—when AI can go from an "assistant" that answers questions to a "digital colleague" that can access data, break down tasks, call tools, and even independently complete a workflow, companies' staffing arrangements start to shift as well.

His career change reveals an important industry signal: in the age of AI agents, new jobs won't necessarily grow out of existing job titles. When tasks once handled by humans begin to be handed to AI, companies also need a group of "intermediaries between humans and AI" to redesign processes, connect data, manage agents, and even teach other employees how to hand their work over to AI.

The corporate org chart is being redrawn amid the generative AI boom. One of the most visible new positions is the Forward Deployed Engineer (FDE).

In Lee's view, the FDE is a cross-domain role that grows out of AI needs. It doesn't require someone to first become a traditional engineer; instead, it starts with understanding business problems, organizing requirements, and assisting with adoption—and requires going into the field to turn AI into a tool that truly integrates into workflows and delivers value.

Ye Ni-shan, another FDE at GNTC with a software engineering background, adds that the team faces different work every week. She first has to understand how different industries operate, then decide which data, permissions, and systems need to be connected—only then does AI come into play.

"Everyone knows about AI, but how exactly do you use it at work?" That is the real challenge companies face.

Lin Tsung-ying, Vice President of Taiwan's largest systems integrator, Systex, observes that what truly holds companies back is often not the lack of models, but that internal data, permissions, and systems are not fully connected. That's why the hot new role of FDE requires going to the front lines of the enterprise to connect data flows, while also understanding employees' original workflows to truly embed AI into daily work.

This brand-new position, which originated at the American software company Palantir, is rapidly rising in popularity amid the AI wave.

AI agents are rewriting not just the front lines but also extending to management: the front lines need FDEs to integrate AI into workflows, while senior levels need someone to decide on AI investment, governance, and priorities. In an IBM survey of 2,000 CEOs this year, 76% of responding companies have already established a Chief AI Officer or a similar management position.

Take Palantir as an example. From its founding, the company served large institutions such as defense and government agencies. Facing highly sensitive data and complex internal systems, simply delivering software was not enough to solve the problem—engineers had to go deep into the field, working with users to complete data integration, system deployment, and troubleshooting. Thus the FDE position gradually took shape.

However, when a company no longer has just a few pieces of software, but hundreds or even thousands of agents that execute work on their own, another role begins to surface: who decides how the entire company uses AI?

In Taiwan, Hotai Motor, with nearly 80 years of history, has gone through the same transformation path.

Last year, Hotai let employees try generative AI tools like ChatGPT and Gemini for free, hoping colleagues would first understand "what AI can do." At the same time, it went further to encourage each department to independently research and find AI use cases in their own work, accumulating more than forty projects within a year.

However, as every department began doing AI, the problem of duplicated investment emerged. "Unit A is doing it, and Units B and C are also doing it—inevitably wasting manpower," said Han Chih-kang, Director of Hotai Motor's MaaS (Mobility as a Service) Advanced Strategy Division, in an interview.

Therefore, last year Hotai established an "AI Development Affairs Bureau" headed by Han as its director. A dedicated team with IT and technical backgrounds helps each unit evaluate tools, develop models, and put reusable functions into an "AI middle platform." Han describes the AI middle platform as a central kitchen: report organization, speech recognition, or marketing modules built by one department can be directly used by others, without each having to redo the work.

Hotai now divides its AI organization into three layers. The bottom layer, "AI Ready," handles data permissions, information security, intellectual property, and governance; "AI Driver" builds the middle platform, models, and tools; and the top layer, "AI Empower," leaves it to each unit to decide which processes to adopt AI for, based on their own business knowledge.

From startup teams to large companies decades old with revenues in the hundreds of billions, demand for generative AI talent is spreading rapidly. FDEs used to mostly appear at AI startups or software companies; now manufacturing is starting to post the same positions.

On 104 Job Bank, Pegatron, Foxconn, and Wistron have all posted openings for "Edge AI Forward Deployed Engineer" (FDE), requiring talent to directly face internal and external customers, understand production line and operational pain points, and then deploy computer vision and large language models into real-world settings, even connecting to existing systems such as MES and ERP.

All of these changes show that the bottleneck for companies adopting AI is shifting from "having a model" to "being able to actually use it." The larger the company, the more complex its internal data, permissions, IT systems, and workflows—and AI experts who only understand models may not truly solve business pain points.

What companies need now are people who understand both technology and the field—people who go to the front lines of business or the factory floor and translate abstract AI capabilities into concrete processes. "AI is a means, not an end," emphasizes serial entrepreneur and WeMo CEO Liu Yu-shun.

This trend has already begun spreading across Taiwan. Across industries such as manufacturing, software, and education, companies are posting more and more FDE positions, with job descriptions generally emphasizing requirements interviews, systems integration, and customer collaboration.

Now that AI has officially moved from the lab into core business processes, will even more new types of jobs emerge in the future? It's worth keeping an eye on.

  • Process integration: What companies truly lack is no longer just people who can build models, but people who can connect AI into existing processes.
  • Co-management: In the future, AI managers will oversee a new type of organizational resource composed of both human employees and AI agents.
  • Knowledge translation: More and more employees may simultaneously play the role of "domain expert + AI process designer." What companies truly need to cultivate will evolve from "knowing how to use AI" to being able to convert one's expertise into workflows that AI can execute.

(Original title: The Hot New AI Job That Dcard and Hotai Are Fighting For! What Is a Forward Deployed Engineer (FDE)?)

Recommended reading:

More 104 [Forward Deployed Engineer FDE] job openings

Don't see a position that interests you?

Curated coverage of finance, international affairs, management, education, The Economist, commentary, and fashion; interactive charts, videos, and other multimedia reports offering in-depth interpretation of global trends and forward-looking ideas. With a gentle, rational attitude and a positive, constructive perspective, we provide a voice that is objective, calm, and trustworthy for Taiwan and Chinese-speaking communities worldwide.