What a Forward Deployed Engineer Is and Why AI Startups Hire Them

A forward deployed engineer (FDE) is a software engineer who works directly inside a customer's business, building and adapting a vendor's product to solve that customer's specific problem instead of staying behind the scenes on the core codebase. The role fuses strong engineering with consulting and hands-on customer discovery, embedding with the client to ship working software in days rather than quarters. It began at Palantir in the late 2000s and has become one of the most sought-after hires at AI companies including OpenAI, Anthropic, Ramp, and Cursor.

What is a forward deployed engineer

A forward deployed engineer is a technical hire who deploys forward, out of headquarters and into the customer's environment, to configure, extend, and integrate a company's product against a real workflow. Unlike a traditional software engineer who ships features for every user at once, an FDE builds for one account at a time, translating messy business requirements into shipped code and feeding what they learn back to the core product team. The job sits at the intersection of engineering, solutions consulting, and product management, and it exists because complex software, especially AI software, rarely works out of the box.

Where the forward deployed engineer role came from

Palantir pioneered the modern FDE. According to Bloomberg reporting, Palantir began working with JPMorgan around 2009, and over the course of the engagement as many as 120 forward deployed engineers were embedded at the bank, working on Metropolis, Palantir's data integration and analytics platform. That approach became the blueprint for how Palantir sells. Send engineers to live with the customer, learn the domain, and build the thing that actually solves the problem. For more than a decade the title stayed mostly a Palantir signature. Then generative AI created a wave of buyers who wanted powerful models but had no clear path to wire them into their own data, security constraints, and processes, and the role broke out across the industry.

Why AI startups are hiring forward deployed engineers now

Foundation models are general. Enterprise problems are specific. The distance between a capable model and a working deployment is exactly the gap a forward deployed engineer closes. An off-the-shelf language model cannot read a bank's internal schemas, respect a hospital's access rules, or fit a logistics operator's exception process on its own. Someone has to sit with the customer, map the workflow, and build the connective tissue that makes the model useful. In the AI era that someone is increasingly an FDE, because the product is only half finished until it meets a real environment.

The demand shows up in the market. OpenAI and Anthropic both post publicly for forward deployed engineers, and startups such as Ramp, Cursor, and Scale AI have hired for the role too. In our own experience building companies, forward deployed engineering is one of the scarcest talent profiles in AI, because the same person needs strong engineering fundamentals, real customer instinct, and the judgment to make product calls in the field. Compensation reflects that scarcity. According to United States Department of Labor H-1B disclosure filings, base salaries for the role commonly sit in the low six figures and run higher for senior hires. At frontier AI labs, total compensation climbs well into the mid six figures once equity is layered on top of base. Those equity-loaded figures are directional rather than a fixed benchmark and vary widely by company, stage, and location, but the direction is clear. Companies are paying a premium for engineers who can make hard technology work inside someone else's building.

What a forward deployed engineer actually does

The work moves through a repeating loop. First comes discovery, where the FDE learns the customer's real problem rather than the one written in the sales deck, a discipline closer to field research than to ticket triage. Strong forward deployed teams treat every deployment as a live experiment in what buyers will actually pay for, which is why the role pairs naturally with a rigorous approach to AI customer discovery. Next comes building, where the engineer integrates the product with the customer's systems, writes the glue code, and prototypes against real data instead of a sanitized demo set. Then comes the feedback loop, where hard-won lessons from one account get generalized and pushed back into the core product so the next deployment is faster and cheaper. Done well, the FDE is not a consultant billing hours. The FDE is a scout who turns one customer's edge case into the whole company's next feature.

Forward deployed engineer vs solutions engineer

The titles overlap, but the center of gravity differs. A solutions engineer or sales engineer mostly supports the sale, running demos, scoping requirements, and answering technical questions to help close a deal. A forward deployed engineer keeps going after the contract is signed and actually builds the deployed system, writing production code inside the customer's stack. The FDE is closer to a founding engineer temporarily assigned to a single account, accountable for whether the software delivers an outcome rather than whether the demo went well. That ownership of the result is what separates the role from adjacent customer-facing jobs.

The forward deployed engineer as a services-to-product wedge

This is where the role becomes strategic rather than tactical, and it is the lens Avante applies as a venture studio that co-founds AI-native companies for Brazil and LATAM. Early enterprise AI revenue almost always looks like services. A team embeds, builds something bespoke, and gets paid for outcomes. The trap is staying there forever, stuck in custom work that never compounds into a product. The forward deployed model is the escape hatch. Each embedded build is designed from the start to harden into reusable software, so bespoke engagements become the raw material for a productized AI copilot instead of a services backlog that resets to zero with every new client.

Avante uses forward deployed engineers as exactly this wedge inside LATAM enterprises. Rather than selling a finished platform into a market that has never bought one, the studio embeds engineers next to the operator, ships a working copilot against a live workflow, and lets real usage reveal what deserves to be generalized. The bespoke work funds the learning. The learning becomes the product. It is a deliberate path from services revenue on day one toward software margins over time, run in a region where enterprise buyers reward proof over promises and a working system in the building beats a polished pitch every time.

When your company needs a forward deployed engineer

Not every startup should hire one early. The role earns its cost when the product is genuinely powerful but genuinely hard to adopt, when a handful of large accounts can move the whole business, and when the founding team is willing to let field lessons rewrite the roadmap. If your product installs itself in an afternoon, you do not need forward deployment. If your buyers need a partner to make the technology real inside their four walls, you do. The same instinct that makes the FDE model work, meeting the customer where they operate and automating their actual process, is the instinct behind any good operator's guide to AI automation. Hire the role when adoption, not capability, is the thing standing between you and revenue.

The bottom line

A forward deployed engineer is the person who makes advanced software real for one customer at a time, and then hands the rest of the company a shortcut to doing it for everyone. Palantir proved the model. AI has made it mainstream. For founders building AI-native companies, especially in markets like Brazil and LATAM where trust is earned in the field, the forward deployed engineer is not a cost of doing business. It is the mechanism that turns bespoke effort into a durable product.

Frequently asked questions

What is a forward deployed engineer?

A forward deployed engineer (FDE) is a software engineer who works directly inside a customer's business, building, integrating, and operating a product against that customer's real data and workflows instead of shipping generic features from headquarters. The role blends engineering, solutions consulting, and product management. It exists because complex software, especially AI software, rarely works out of the box, so someone has to embed with the customer, learn the workflow, and make the technology deliver an outcome.

Which companies hire forward deployed engineers?

Palantir pioneered the role. Today OpenAI and Anthropic both post publicly for forward deployed engineers, and startups such as Ramp, Cursor, and Scale AI have hired for it too. Beyond the frontier labs, any company selling powerful but hard-to-adopt software uses the model, and venture studios use forward deployed engineers to put AI to work inside enterprise customers.

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

A solutions engineer or sales engineer mostly supports the sale by running demos, scoping requirements, and answering technical questions to help close a deal. A forward deployed engineer keeps going after the contract is signed and builds the deployed system, writing production code inside the customer's stack. The FDE is closer to a founding engineer temporarily assigned to a single account and is accountable for whether the software delivers an outcome, not whether the demo went well.

How much does a forward deployed engineer make?

United States Department of Labor H-1B disclosure filings indexed by h1bdata.info show a median base salary near 155,000 dollars across 196 forward deployed engineer roles, with Palantir accounting for 127 of them. At frontier AI labs, total compensation climbs well into the mid six figures once equity is layered on top of base. Those equity-loaded figures are directional and vary widely by company, stage, and location.