Forward Deployed Engineer: AI's Hottest New Job, and What It Actually DoesForward Deployed Engineer：AI 时代的新宠岗位，到底干什么？

An "arms race" for AI jobs

First, a look at some recent news in the AI world about a new job title: Forward Deployed Engineer (FDE).

Google is doubling down on the FDE role and has drastically simplified its interview process. Google Cloud CEO Thomas Kurian announced that they are forming a brand-new, AI-centered division under the Go-To-Market team, and are aggressively recruiting FDEs for it.

Reportedly, their interview process has been dramatically compressed — from what used to take several weeks and 4-6 rounds, down to just two rounds within two days. It seems Google isn't just eager to fill these openings — it's practically desperate.

This past Monday (May 11), OpenAI announced the formation of "The OpenAI Deployment Company." It's an independent entity backed by $4 billion in private equity funding, valued at $14 billion, with investors including TPG, Advent, and others. It appears OpenAI itself isn't a direct investor, but rather plays the role of a partner.

The announcement specifically mentioned FDEs, stating that their job is to "work closely with business leaders, operators, and frontline teams to pinpoint where AI can create the most value, and redesign organizations' underlying infrastructure and critical workflows around AI, ultimately converting those gains into durable, lasting systems."

This shows that FDEs will play an extremely critical role in OpenAI's enterprise sales business — their task is to ensure the company's AI systems actually work in customers' real-world operations and genuinely create value. By outsourcing this business to the newly formed "Deployment Company," OpenAI can also free itself up to focus on developing more powerful AI models, while leaving the tedious work of customer integration to its partners and their FDEs.

A related development: OpenAI acquired Tomoro. This is a UK-headquartered AI company founded in 2023, with 150 FDEs across the UK, Asia, and Australia. This is also the first acquisition since the founding of the "OpenAI Deployment Company."

Anthropic is following the same playbook, setting up its own independent FDE consulting company. Last Monday (May 4), Anthropic released an extremely vague announcement revealing this new business, without even disclosing a name, and with scant details on the investment.

Known investors include Anthropic itself, Blackstone, Hellman & Friedman, and Goldman Sachs. The new company's mission is to "partner with mid-market companies across industries to bring the large language model Claude into their most critical business operations."

Anthropic's calculations seem to mirror OpenAI's exactly: bring in outside capital to set up an independent company, and have the FDEs inside it help businesses integrate Claude into their systems. It's easy to predict that, as a result, the amount of Claude tokens these companies purchase will hit an all-time high.

Explaining in plain language what an FDE actually is

So what exactly is an FDE? The full name is Forward Deployed Engineer, abbreviated FDE. Translated literally, this means "engineer deployed at the front line," but the name alone doesn't make it easy to understand what the role actually does.

In one sentence: an engineer stationed on-site at the customer's company, writing code.

In more detail, this role sits somewhere between a software engineer, a solutions architect, and a consultant — but is more hands-on. They sit directly inside the customer's company, using their own company's AI technology to solve real problems for the customer.

You might ask, isn't this just a consultant? Not quite. A consultant typically hands you a slide deck telling you "here's the best way to do it"; an FDE hands you code and helps you actually get it done. A solutions architect generally draws architecture diagrams and writes technical proposals; an FDE does all that, plus gets hands-on writing code, wiring up interfaces, and debugging on-site.

To put a rough ratio on it: about 25% writing code, 50% integration and debugging, 25% meetings and communication. In practice, the time spent quietly writing code is probably even less than that.

Actually, Palantir was the original pioneer

When it comes to FDEs, this isn't something that just emerged in the AI era — it's a playbook Palantir already had down pat back in the 2010s.

Palantir builds data analytics platforms, and its early customers were almost entirely the US military and intelligence agencies, whose needs were all classified and simply couldn't be communicated through conventional channels. So Palantir simply sent engineers to be stationed on-site with customers, to observe their needs up close and iterate rapidly in the field.

These on-site engineers (Palantir calls them "Deltas") did more than just deliver projects — they had an even more important task: distilling common requirements from the customer side and feeding them back to the product team to be built into standardized features.

By 2016, Palantir already had more FDEs than regular engineers, truly defining the role.

Three companies betting on FDEs, three different paths

OpenAI is the most aggressive. It formed the OpenAI Deployment Company, with TPG, McKinsey, Bain, and Capgemini all on board, pushing the valuation as high as $14 billion, and directly acquired a UK company, bringing 150 ready-to-work FDEs on board immediately. It promises a minimum return rate of 17.5%, more like an infrastructure investment.

Anthropic is a bit steadier. It teamed up with Wall Street giants like Blackstone, Goldman Sachs, and Apollo to form a joint venture, with an initial investment of $1.5 billion, primarily targeting the mid-market. These investors hold stakes in a huge number of companies, which naturally makes them the ideal user pool for the Claude model.

Google is the most traditional. It's hiring people directly, with FDE roles distributed globally, and the pay isn't low either — senior-level total compensation in the US can exceed $400,000. But the biggest difference is that Google's FDEs receive Google stock, whereas OpenAI's and Anthropic's FDEs work at independent companies, with no direct stake in the parent company's fortunes.

Translating the "real talk" behind Google's FDE job posting

Corporate job postings are the kind of thing that often leave people scratching their heads — let's translate it:

Although this might sound a bit like griping, in reality every company's job description is similar. Being mentally prepared for this makes it easier to know whether the role is actually a fit for you.

The soul-searching question: is an FDE still basically a consultant?

Let's look at it along three dimensions.

First, organizational affiliation. Palantir's FDEs belong to the product team, and rise and fall together with the parent company. But OpenAI's and Anthropic's FDEs belong to independent companies, which takes a toll on information flow, sense of identity, and career development paths.

Second, the feedback loop. An FDE's greatest value is discovering customer needs and feeding them back into the product. But with an organizational gap separating the independent company from the parent company, this feedback channel may get blocked, and the FDE risks becoming little more than a "consultant who happens to write code."

Third, alignment of interests. Google's FDEs hold parent-company stock, so their interests are aligned. OpenAI's and Anthropic's FDEs instead get a stake in the independent company's returns — even if the parent company's valuation soars to the sky, none of that belongs to them.

The conclusion is that OpenAI's and Anthropic's FDEs are already closer to consultants, while Google's is closer to the traditional FDE model.

Who should be paying attention to FDEs?

Break it down into three groups:

New graduates: an excellent opportunity. Software roles at big tech companies are shrinking, but FDE hiring is booming — you can quickly gain exposure to enterprise-grade AI projects and grow faster.

Senior engineers: might feel like a "step down" — customers rotate frequently, and there's a lack of long-term belonging; but if you're thinking about starting your own company or want to get closer to the business side, an FDE role is an excellent window into deep enterprise needs.

Non-technical backgrounds: the bar is still quite high — it's not something a few months of learning Python will get you through.

The competition in the AI industry has quietly shifted

Over the past three years, the AI industry has been competing on model size and benchmark scores. Now the question has changed — most enterprises aren't short on models, they're short on people to help plug those models into their business.

OpenAI opened with $4 billion, Anthropic put in $1.5 billion, and Google compressed its hiring process down to two days. These massive investments show that the way AI companies make money has changed — from selling models to selling deployment.

Put more broadly: for every dollar spent training a model, another dollar may need to be spent actually getting that model to run in practice.

And FDEs happen to be standing right at the forefront of this turning point.