A forward deployed engineer is a software engineer who works inside the customer’s team and systems, and stays until the software does real work there. The job title is suddenly everywhere. This guide explains what the role is, where it comes from, why AI made it spread, and what it means for a company that wants AI to do more than answer questions.
There are two kinds of reader for this topic. Some want the job. Most of this guide is for the other kind: the leader who keeps hearing the term and wonders whether the company needs one.
Key takeaways
- A forward deployed engineer, or FDE, builds inside your systems, with your people, and is measured on whether the result works.
- The role is usually traced to Palantir. Its Chief Technology Officer says he coined the term in 2007.[1]
- AI made the role spread, because AI products need deep connections to a company’s own systems before they are useful.
- An engineer alone is half of the answer. Someone also has to train your people to use what was built.
What is a forward deployed engineer?
A forward deployed engineer is an engineer who is placed with a customer instead of staying in the vendor’s office. Three things define the role:
- They work where the customer works. In the customer’s own systems, on the customer’s own data, next to the people who do the job.
- They write real software. Not slides and not a demo. Code that runs in production, which means in daily use.
- They are measured on the outcome. The question is not “did we deliver the hours?” It is “does it work on Monday?”
Palantir puts it in one sentence in its own job description: “Forward Deployed Engineering is a radical commitment to the outcome.”[2]
Salesforce describes its own FDEs like this: “They code, consult, and translate agentic AI into working solutions, often while sitting side by side with the customer.”[3]
Where the role comes from
The term is usually traced to Palantir. Shyam Sankar, the company’s Chief Technology Officer, says he coined it in 2007. According to The Register, the idea came from a conversation about French restaurants. Sankar recalls Palantir’s CEO Alex Karp saying that in a good restaurant “the wait staff is actually part of the kitchen staff. They intimately understand the food, the methodology, and the technique.”[1] In Sankar’s account, he was asked to build the same thing for engineering: people who serve the customer and understand the kitchen.
The role is demanding. Palantir’s own posting says to expect 25 to 50% travel.[2]
Why AI made the role spread
Then AI products arrived, and many companies found the same problem: the tool was bought, and nothing changed.
The job market reacted fast. Monthly job listings for the role rose by more than 800% between January and September 2025, according to an analysis of Indeed data reported by the Financial Times.[4] That growth started from a very small number. It shows a direction, not the size of the job market. Salesforce says it tripled its own FDE team in six months.[3]
Why now? The investor Andreessen Horowitz gives a simple picture: “Enterprises buying AI are like your grandma getting an iPhone: they want to use it, but they need you to set it up.”[5] In its words, enterprise AI products “require deep integrations and context”. An AI tool that cannot see your CRM (the customer database), your documents and your rules can only chat. Someone has to connect it, safely, to the real work.
What does a forward deployed engineer do in a week?
The details differ by company. For example, a week inside a customer team might look like this:
- Watch the work. Sit with the people who do the job. See the list that is exported every Monday and pasted into three systems.
- Build. Connect the AI to the systems where that work happens, with read-only access first.
- Test with the team. Put the first version in front of the people who will use it. They are quick to say what is wrong.
- Go live. Move it into daily use, inside the tools the team already has.
- Review. Look at what was used and what was not, and decide what comes next.
The person who hears the problem is the person who fixes it.
Forward deployed engineer, consultant or solutions engineer?
The three roles are easy to confuse, because all three talk to customers. They do different jobs.
| Forward deployed engineer | Consultant | Solutions engineer | |
|---|---|---|---|
| Main job | Make it work in the customer’s systems | Advise on what to do | Show what the product can do |
| Writes production code | Yes | Usually not | Usually not |
| When | After the decision, until it runs | Mostly before the decision | Mostly before the sale |
| Typical goal | A working result | Good advice | Supporting the sale |
If you are comparing outside help in general, our guide to choosing an AI consultant has seven questions to ask.
The other half: the business operator
An engineer can build a tool that works perfectly and that nobody uses. This is one way for AI projects to disappoint, and more engineering does not fix it.
BCG has a rule of thumb for this. About 10% of the value of AI comes from the AI application itself, about 20% from data and technology, and about 70% from how people and processes change.[6] It is one consulting firm’s framework and not a measurement, but it points at the missing half.
That is why we work in pairs. One builds. One trains your people.
- The engineer builds the use cases in your tools, on your data, in your own environment. Real systems, not a demo.
- The business operator trains your people while we build. Every week there are office hours, where people bring their real work.
The engineer makes it run. The operator trains your people to use it. See how the weekly office hours work.
What it looks like in our own work
I spent more than twenty years in software go-to-market before PromptPartner, and helped grow WebPros (cPanel, Plesk, WHMCS) to a €1.5 billion exit to CVC. The lesson I took from those years: a tool only works when the people who build it sit close to the people who use it.
So we work close to the people who use the tool. For example, at a European software company above €100 million in annual recurring revenue, we built inside the client’s own environment, with their people, from the first day. We shipped five working tools into the systems the team already used. One of them turns a week of work on sales material into minutes.
Because the work lives in the client’s own systems, there is no handover at the end. There is nothing on our side to hand over. Read the case study.
When does your company need one?
Not every company does. Three signs suggest that you do:
- You bought AI licences and little changed. People use the chat. The work is the same.
- The work crosses several systems. The customer database, email, the calendar. A tool that lives inside only one of them cannot follow the work.
- Your own engineers have no time. They keep the product running. Nobody is free to connect AI to the business side.
If you already know exactly what you want built and it sits in one system, an agency with a clear brief may be enough. If you do not yet know which work is worth doing with AI, start by measuring. Our method starts with a four-week assessment, and this page shows what agents then do in your systems.
Frequently asked questions
What are the skills of a forward deployed engineer?
Strong software engineering is the base. On top of it come the skills that are harder to find. Salesforce names “judgment, pattern recognition, and the ability to deliver hard truths to C-suite stakeholders without alienating them”. C-suite means top management.
Who coined the term forward deployed engineer?
The term is usually traced to Palantir. Its Chief Technology Officer, Shyam Sankar, says he coined it in 2007 after a conversation with CEO Alex Karp about why the waiters in French restaurants are part of the kitchen team.
Is a forward deployed engineer a consultant?
Not only. A consultant mainly advises. A forward deployed engineer also builds software that runs in the customer’s systems, and is measured on whether it works.
Why are forward deployed engineers in demand?
Because AI products need deep connections to a company’s own systems and data before they are useful. Job listings for the role rose by more than 800% between January and September 2025, according to an analysis reported by the Financial Times, from a small starting number.
Does my company need a forward deployed engineer?
Perhaps, if you bought AI licences and little changed, if the work crosses several systems, and if your own engineers have no time. An engineer alone is half of the answer. Someone also has to train your people.
How do I become a forward deployed engineer?
Build things that work in real companies, and learn the business side as well as the code.
Looking for the job and not the service? See our open role.
Next step
If you want to know what an engineer and an operator inside your team would start with, book a 30-minute call.
No pitch. You’ll leave with at least one recommendation you can use — whether or not we ever work together.
Sources
- [1]The Register, on the origin of the forward deployed engineer, 3 October 2026— reports Shyam Sankar’s own account
[2] Palantir, job description “Forward Deployed Software Engineer” — retrieved 5 October 2026
[3] Salesforce News, on forward deployed engineers, 18 March 2026 — a vendor’s own account
[4] PYMNTS, 10 March 2026, citing the Financial Times and Indeed data — a re-report of the original analysis
[5] Andreessen Horowitz, “Services-led growth,” 4 June 2025 — an investor’s view
[6] BCG, Artificial Intelligence — the 10-20-70 rule of thumb
