Table of Content

Read time

8 minutes

What is a forward deployed engineer?

A forward deployed engineer, often shortened to FDE, is a software engineer who works inside a customer’s own environment to build and run the solution that makes a product or a platform actually work there. Unlike a consultant who recommends, or a support engineer who answers tickets, an FDE writes production code against the customer’s real systems, stays through the rollout and carries what they learn back to the product team. The role began at Palantir and has spread to AI labs, defense technology and enterprise software.

TL;DR

  • A forward deployed engineer is a hands-on engineer embedded with a customer. They configure, integrate and often build the solution on the customer systems, and they stay accountable until it works.
  • The role is usually traced to Palantir, where it is reported to date to the early 2010s and was known internally as a “Delta”. Sources disagree on the exact year, so treat any single date with care.
  • Demand has jumped. Trade press reports FDE job postings growing several hundred percent through 2025, and an analysis of about 1,000 postings found three different jobs hiding behind one title.
  • You need an FDE when integration is complex or unpredictable, the customer is large and slow-moving, or the first deployment has to prove the product works. You probably do not need one for a standard, well-documented setup.
  • Your options are to hire one, bring in a contractor or use a partner who embeds a team. A worked example below compares an in-house hire with a partner for a six-month integration, using placeholder numbers.
  • Salary and skill figures come from a third-party analysis of public postings and are medians and shares, not guarantees. The cost figures in our worked example are placeholders.
  • CommercePundit embeds engineers with client teams to connect systems and build automation. See the last section.

What does a forward deployed engineer actually do?

The simplest way to understand the role is to follow one through a week. Monday might be spent on site, or on a call with the customer, watching how an operations team really moves data between an ERP, a spreadsheet and an email inbox. Tuesday is code: an integration, a data pipeline or an AI workflow written against the customer’s actual systems, with their messy field names and their access rules. Wednesday is debugging something that only fails with production data. By Friday the engineer is writing up a gap in the underlying product and sending it to the people who can fix it.

The Pragmatic Engineer describes the split in time as three areas: customer project work, which is the main focus, improvements to the company platform when it blocks a customer, and the usual internal work such as reviews and meetings. That second area is what separates an FDE from a pure services engineer. Field experience flows back into the product, so the next customer gets a smoother start.

Palantir, which is credited with creating the role, framed the difference neatly. A traditional developer works on one capability for many customers, while a forward deployed engineer works on one customer with many capabilities. That framing still holds: the FDE is measured on whether this customer succeeds, not on shipping a feature to everyone.

Where did the role come from, and why is everyone hiring for it now?

Most accounts trace the role to Palantir. The Pragmatic Engineer places its creation in the early 2010s and says engineers carried the internal title Delta, and that until around 2016 Palantir employed more forward deployed engineers than ordinary software engineers. Other write-ups date it earlier, to the mid-2000s or to Palantir’s founding, so the honest answer is that the exact year is disputed. What is not disputed is the direction: the model has moved well beyond Palantir.

The reason is mostly AI. A language model API is easy to call and hard to put to work inside a company with legacy systems, strict permissions and unclear processes. Vendors found that the gap between a promising demo and a production deployment is filled by engineers who can sit with the customer, learn the process and write the glue. Reports citing Financial Times and Indeed data put the growth in FDE postings at more than 800% between January and September 2025, and Bloomberry’s analysis of 1,000 postings found a year-over-year rise above 1,000%. Those are trade-press and vendor figures that overlap and are often repeated from one another, so read them as a strong direction, not a precise count.

The Bloomberry analysis is useful for another reason. It found that the title covers three different jobs. Around 60% of the postings were builder roles, engineers who embed with customers and run production systems. About 30% were sales engineer roles with a new title, running demos and proofs of concept, often with a quota. The remaining 10% or so were internal tools roles building automation for the company’s own teams. If you are hiring, or buying the service, that distinction matters more than the title.

How is an FDE different from a solutions engineer, a consultant or an agency?

These roles overlap, and the boundaries blur in practice. The clearest differences show up in who writes code and how long they stay.

Role Writes production code on the customer side? Typical engagement Feeds the product?
Forward deployed engineer Yes Long-term, embedded with one customer Yes, usually a core part of the job
Solutions engineer or architect Rarely; prototypes and demos Pre-sales and advisory Sometimes, as feedback
Consultant Sometimes Project with a defined recommendation No
Agency or delivery partner Yes Fixed scope or time-and-materials project Not their product; they feed your delivery process

Notice that a good delivery partner can fill the same seat as an FDE for a customer who does not have the product company behind them. That is the point of the next section.

When do you actually need one?

You need forward deployed engineering when the cost of a failed integration is higher than the cost of embedding someone. Four situations make that true.

First, when you are rolling out something that touches several existing systems and nobody is sure how they will behave together. AI projects over ERP data are the typical case, because the model is the easy part and the permissions, data quality and edge cases are the hard part. Our guides to AI in construction and AI agents in ERP are both built around that point.

Second, when the customer is large and slow. A startup-minded engineer sitting inside a bureaucracy can push an integration forward faster than the customer can alone. Third, when the first deployment is also the sales proof, and the buyer wants to see the thing working on their data before they commit. Fourth, when customer feedback should shape the product roadmap, which is the original reason Palantir built the model.

You probably do not need one when the setup is standard and documented, when the work is a one-off configuration, or when you only need advice. In those cases a good consultant, or simply better documentation, is cheaper.

Hire one, contract one or use a partner? A worked example

Take a six-month goal: connect an ERP, a field system and an AI assistant so staff can ask questions of their own data. These numbers are placeholders for illustration only. For the in-house option, use a base salary of $174,000, which is the median base Bloomberry reported across postings that disclosed pay. Add a loaded cost of 30% on top for benefits and overhead, a recruiting fee of 20% of base and about two months to hire. For the partner option, assume $150 an hour for 120 hours a month.

The in-house route costs about $18,850 a month once the person starts, so six months of project work is about $113,100, plus $34,800 to recruit, for a total near $147,900. The first line of code, though, is written about two months after you decide, so the work finishes around month 8. The partner route costs about $18,000 a month, or $108,000 over the six months, starts in the first month and finishes at month six, with no one left on payroll afterward.

On these inputs the partner is cheaper by about $39,900 and about two months faster for a one-time integration. Flip the assumptions and the answer flips with them. If the work is continuous for years, or the person will also be your product engineer, hiring wins. The decision turns on how long you need the capability and how fast you need it, which is why the numbers belong in your own spreadsheet, not ours.

What should you look for when you hire or buy forward deployed engineering?

Look first for someone who writes code and talks to the customer in the same hour. The Bloomberry analysis found Python in about two thirds of builder postings and TypeScript in about a third, alongside cloud platforms such as AWS, and AI agents or large language model experience in roughly a third. It also found that around 68% of the roles required travel. Those are the skills of a person who can sit in a meeting in the morning and ship a fix in the afternoon.

Look for evidence of ownership. Ask for a deployment they carried from first conversation to production, what broke, and what they changed in the product afterward. Ask how they handle permissions and data access, because that is where most integrations stall. If you are buying from a partner, ask who will actually be embedded, how many other projects they will split time across and what you own at the end: repositories, documentation and the infrastructure accounts. Our guides to staff augmentation, hiring dedicated developers and hiring AI developers cover the contract side.

One more check is whether they work from specs and keep the work reviewable, especially if AI coding agents are involved. Embedded engineers move fast, and fast without discipline leaves a mess for your team. We describe the approach we use in our guide to spec-driven development.

What does this look like in our own work?

We do forward deployed style work for contractors, ecommerce brands and operations teams: sit with the people who run a process, connect the systems they already use and build the automation on top. In our ArchetypeDG project we built a custom commerce platform with agents working under written specs, which you can read in the ArchetypeDG case study. In the CMES project we connected ERP, field and HR data and added assistants that read it without changing it, described in the CMES case study. If you want an embedded team for an AI or integration project, our AI development services and guide to AI agent development companies are good places to start.

Frequently Asked Questions

What is a forward deployed engineer?

A forward deployed engineer is a software engineer embedded with a customer who builds, configures and runs a solution inside the customer environment. They write production code, stay through the rollout and feed what they learn back to the product team.

What does FDE stand for?

FDE stands for forward deployed engineer. You will also see FDSE for forward deployed software engineer, and Palantir used the internal name Delta for its version of the role.

Who invented the forward deployed engineer role?

Palantir is widely credited with creating it. Sources disagree on when: the Pragmatic Engineer places it in the early 2010s, while other write-ups date it to the mid-2000s or Palantir founding, so no single year is confirmed.

What is the difference between a forward deployed engineer and a solutions engineer?

A solutions engineer is mostly advisory and pre-sales, building demos and prototypes, often with sample data. A forward deployed engineer writes production code on the customer systems, stays with the customer long term and usually contributes to the product.

Is a forward deployed engineer the same as a consultant?

No. A consultant typically gives recommendations for a defined project. A forward deployed engineer stays with the customer, writes code and is accountable for the solution working in production.

How much does a forward deployed engineer earn?

An analysis of about 1,000 public postings by Bloomberry put the median base salary near $174,000 across postings that disclosed pay, with builder roles ranging roughly from $140,000 to $250,000. Treat these as medians from one analysis, not a guarantee.

What skills does a forward deployed engineer need?

Strong software engineering, most often Python and TypeScript, cloud experience, increasingly AI agent and large language model work, and the ability to talk to non-technical customers. The same analysis found travel was required in about two thirds of roles.

When should a company hire a forward deployed engineer?

When integration is complex or unpredictable, the customer is large and slow-moving, the first deployment has to prove the product works, or customer feedback should shape the roadmap. For standard, documented setups it is usually unnecessary.

Can a partner act as a forward deployed engineer?

Yes. A delivery partner can embed engineers with your team for a defined period, which is often faster and cheaper for a one-time integration. Check who is embedded, how much time they give you and what you own at the end.

Why are forward deployed engineers in demand now?

Mostly because of AI. Models are easy to call but hard to put to work inside companies with legacy systems and strict permissions, and FDEs fill the gap between a demo and production. Trade-press reports show job postings growing several hundred percent in 2025.

Bottom Line

A forward deployed engineer is the person who makes a product or a platform work inside a customer’s real environment, and who carries the lessons back. Decide whether you need one by looking at integration risk and how long you need the capability, and compare hiring, contracting and a partner with your own numbers. If you want an embedded team to connect your systems and build AI automation on top, see our AI development services and custom web development services.

Hardik Surana
Customer Success Manager

As a Customer Success Manager, I specialize in bridging the gap between structured execution and long-term strategic growth. Drawing on my extensive background in traditional project management and account oversight, I ensure our clients don't just hit milestones, but achieve measurable business outcomes. My approach is rooted in providing a seamless transition from implementation to partnership, leveraging my technical discipline to anticipate challenges before they arise. I am committed to acting as a dedicated advocate for my customers, turning complex requirements into streamlined paths for success. By blending operational rigor with a relationship-first mindset, I focus on driving the continuous value that keeps our partners thriving.

What our clients says

An assemblage of our most passionately crafted works alongside forward-thinking clients and friends throughout the years.

Eric Truong
Eric Truong CEO, LA Nails Supply
" Commerce Pundit’s collaboration created a stunning website representing our beauty nails business. Their SEO strategies boosted rankings and traffic.Transparent communication and prompt feedback incorporation exceeded expectations. "

0%

Increase in orders

0%

Increase in revenue

0%

Increase in site traffic

Contact Us