What is FDE? Why Does the Software Industry Need FDE?

Over the past two years, the software engineer job market has contracted overall, yet FDE (Forward Deployed Engineer) positions have grown remarkably fast compared to other roles. This trend has sparked numerous discussions in the community about what FDEs are and why they matter.

These conversations revolve around several key questions: What exactly is an FDE role? How does it differ from a traditional software engineer (SWE)? Why is the software industry embracing FDE positions? What kind of people are suited for FDE work? And if someone wants to transition from SWE to FDE, what additional skills do they need to develop?

In this article, we'll explore these questions and understand this emerging trend in the software industry.

What is FDE?

Literally speaking, the "F" in FDE stands for "Forward," meaning these engineers need to be at the front lines where customers operate. The "D" stands for "Deployed," indicating they don't just visit the front line—they actively execute tasks there. The term actually comes from military language: "Forward Deployed Unit" refers to military units stationed on the front lines carrying out operations.

In military operations, there's a central command that sets overall strategy and produces military supplies, while forward deployed units take those strategies to the field and execute actual missions. The software industry has a similar structure. Alongside R&D (Research and Development) teams, there are forward deployed engineers who bring research outcomes to customers' environments, making customizations to solve their specific problems.

Consider OpenAI, which has been recruiting FDEs in large numbers. The core product comes from the R&D team's AI models, while FDEs help OpenAI's enterprise customers actually implement these models. This includes designing the best solution for each customer's needs, developing a prototype to prove the approach works (so customers feel confident buying in), and then continuously supporting customers as they transition from prototype to production.

In a previous interview (link), Colin Jarvis, OpenAI's FDE team lead, shared how when ChatGPT's API first launched, the FDE team developed a customized solution to convince Morgan Stanley to adopt it. They created a system that retrieves information from large document collections and uses ChatGPT's API to synthesize insights valuable for wealth advisors.

In this partnership, the FDE's role was to be on the front lines, understanding what pain points wealth advisors faced in their daily work, understanding why they were hesitant to use AI even when the models performed well, and then building a customized solution that addressed those concerns and earned their trust.

Breaking down the division of labor: OpenAI's researchers train the AI models, the product team converts those models into API products, and then FDEs build customized solutions on top of those APIs to win customers' trust and adoption.

How is FDE Different from SWE?

After reading the above, you might wonder: how does FDE differ from SWE (Software Engineer)?

After reviewing over 50 FDE job descriptions, we believe FDE is fundamentally a software engineer role, but represents a specialized subcategory within the broader engineering field. Just as there are Frontend Engineers, Backend Engineers, and Infrastructure Engineers within software engineering—all of whom are still software engineers—FDE is another specialized variant of the same discipline.

Like other software engineer specializations, FDEs have their particular focus. To understand this, we need to look at how the FDE role was originally created.

Why Did the FDE Role Emerge?

One of the most well-known origins of the FDE role comes from Palantir. Palantir's customers were primarily in intelligence and defense sectors, each with distinct needs. Rather than serving all customers with a single product, the company needed to customize solutions for different departments.

In this context, the dominant software model of the past two decades became impractical. The mainstream business model has been one software product serving all customers—think of the B2B products teams use (Google Workspace, Linear) or the B2C platforms we use daily (social media, food delivery apps). Everyone uses the same software product regardless of individual variations in their specific needs.

This approach embodies the essence of software abstraction: by abstracting common patterns, one product serves all users. For instance, whether user A wants to order beverages or user B wants to order snacks, both needs fit within UberEats' abstraction—no customization required. Most companies therefore hire general software engineers, or specialists in frontend and backend product development.

However, because Palantir's customers had highly customized requirements, the company needed people beyond the core product team to develop custom features for different clients. So Palantir recruited general software engineers for core product development, and separately hired FDEs to help customers build customized solutions.

While FDEs' primary task is to be at the customer's front line, the role carries another important responsibility: bringing observations from the field back to Palantir's core product team, helping them iterate on the core product itself. More precisely, FDEs operate on the front lines but also interface with both customers and the core product team.