Editor's Note: As companies like OpenAI and Anthropic start building AI Forward Deployed Engineer (FDE) teams, a role originating from Palantir is gaining popularity in Silicon Valley. The core value of FDE is to go on-site with clients and transform general large models into Agent workflows tailored to specific business processes.
However, what this article truly discusses is not just the new FDE role, but how job structures in the AI era are being reshaped. The author believes that, compared to a small number of FDEs deployed on-site at clients to serve the implementation of a specific vendor's product, the future demand will be greater for in-house AI Engineers within enterprises. These individuals need to understand cues, Agent frameworks, evaluation systems, and also be able to use AI programming tools such as Claude Code and Codex to truly embed AI capabilities into software and business systems.
This also means that AI's impact on the job market may not simply be "substitution." It is more likely to first create a batch of new generalist positions, and then, similar to the differentiation of software engineers into front-end, back-end, mobile, and DevOps in the past, further evolve into more specialized roles such as LLMOps, Evals Engineer, AI Data Engineer, and so on. The truly scarce individuals will be those who not only understand engineering implementation but also grasp the business context.
Below is the original article:
A new and highly anticipated role has emerged in Silicon Valley recently: the AI Forward Deployed Engineer (FDE). These engineers are embedded within client organizations to help customize solutions, such as building and fine-tuning Agent workflows tailored to the client's specific needs. Since OpenAI and Anthropic began assembling new teams and deploying FDEs to client organizations, I have also heard many people starting to pay more attention to the FDE career path.
The rise of the FDE role is being driven by AI workloads and is an example of AI creating new jobs. This also demonstrates that the narrative of an imminent "jobpocalypse" where the job market is expected to collapse is unfounded—there will still be a large number of AI and non-AI related roles in the future. However, as explained in the following text, I believe the number of AI Engineer positions will far exceed that of FDEs.
The FDE role was pioneered by Palantir approximately twenty years ago. At that time, Palantir would send engineers to government agencies to work on-site in a secure, air-gapped environment. In addition to strong technical skills, an FDE also needs to have good communication skills and sometimes requires a certain level of business acumen. For example, they may need to communicate with clients to understand their needs, prioritize project strategies, explain complex technology, and respectfully but firmly provide feedback when clients make unrealistic demands. Today, the FDE role is once again in the spotlight, mainly because the process of embedding a off-the-shelf large language model into enterprise operations and transforming it into a customized Agent workflow tailored to specific business needs requires a significant amount of practical implementation work.
However, I believe that the scale of AI engineer positions will be much larger. While a company may be willing to accept a few FDEs for internal collaboration, most companies would prefer to involve more of their own employees in project development. In the case of my organization, for instance, we do hire FDEs, but the number of AI engineers we hire is much greater. Additionally, a common concern among clients is the difficulty in finding truly "vendor-neutral" FDEs. After all, the FDE's task is essentially to deeply integrate a vendor's product into the enterprise system. At this stage, it is difficult to predict which AI service will be the best choice a year from now, so "optionality" is crucial, meaning that a company should be able to choose the most suitable vendor in the future. In contrast, if an FDE deeply integrates a company's business processes with a particular vendor, it significantly limits this optionality.
Currently, I see a rapid increase in demand for AI engineers in the market. These engineers can build applications using AI software components such as LLM prompts, Agent frameworks, evaluation systems, and can efficiently use AI programming Agents such as Claude Code, Codex, Antigravity CLI, and OpenCode. As the role of AI engineer matures, I expect it to further split into more specialized positions. Similar to the generic "software engineer" role decades ago, which later diversified into frontend, backend, mobile, data engineering, DevOps, and other directions.
What specialized AI engineering positions will emerge in the future? I cannot be certain yet. There may be AI FDEs, LLMOps engineers, evaluation engineers, AI data engineers, Harness engineers, and some new positions that we have not named yet. But at least for now, many generalist AI engineers are already creating tremendous value. Outstanding AI engineers are in high demand. As this field continues to mature over the next decade, I also expect to see more professional specialization within AI engineering, creating more new job opportunities as a result.
Welcome to join the official BlockBeats community:
Telegram Subscription Group: https://t.me/theblockbeats
Telegram Discussion Group: https://t.me/BlockBeats_App
Official Twitter Account: https://twitter.com/BlockBeatsAsia
