A Forward Deployed Engineer (FDE) is an engineer embedded with a client’s team who builds a solution and takes it to production directly in the client’s environment: its data, its systems, its constraints. The role comes from Palantir; in 2026 it became a leading model for enterprise AI deployment, picked up by OpenAI, Anthropic, Google Cloud, AWS and Microsoft.
Where does the Forward Deployed Engineer come from?
The term comes from US military vocabulary: being “forward-deployed,” close to the front lines. Palantir has used it since at least 2009 for the engineers it places with its clients. Their job: make the software work in the client’s real-world environment, and feed what they learn back into the product.
Generative AI took the model mainstream. Many AI pilots stall at integration rather than at the model: scattered data, access rights, legacy systems, team adoption. In 2026, vendors scaled up: Anthropic, OpenAI (The OpenAI Deployment Company), Google Cloud, AWS and Microsoft all announced dedicated organizations for embedding engineers with customers.
What does an FDE actually do?
- Starts from the business process, not the technology: what work should AI prepare, check or automate?
- Connects AI to existing systems: ERP (SAP, Priority, Oracle), Teamcenter PLM, CRM, documents, business applications.
- Builds and delivers in small iterations, with the users.
- Stays until production and adoption: a solution that is deployed but unused creates no value.
- Transfers knowledge: documentation, training and handover so your teams can run the solution.
FDE, consultant, staff augmentation: what’s the difference?
| Forward Deployed Engineer | Consultant | Staff augmentation | |
|---|---|---|---|
| Goal | A solution in production and in use | A diagnosis, a recommendation | Engineer time billed by the day |
| Deliverable | Code running in your systems | A report, a roadmap | Whatever the client directs |
| Position | Inside your teams, committed to a result | Alongside the teams | Under the client’s direction |
When should you bring in FDEs?
- An AI pilot works as a demo but never reaches production.
- AI needs to read from or write to your systems (ERP, PLM, CRM) rather than stay an isolated tool.
- Your teams lack the time or skills to integrate, secure and maintain the solution.
- You want to move fast, with clear governance: access rights, traceability, human validation.
How Tocndix applies the FDE model
Our engineers work embedded with our clients’ business and IT teams, on site or remotely. They combine AI, Mendix low-code and custom development, with rare expertise in Siemens Teamcenter. Our engagements are fixed-price, scoped around a production outcome. Learn about our FDE model.
Frequently asked questions
Is an FDE a developer?
Yes, and more: an FDE writes code, and also spends a lot of time on integration, data and working with users.
Do you need a full-time FDE?
No. The engagement scales to the project: one engineer or a small team, full-time or part-time, until the solution is in production.
Can an FDE work remotely?
Yes. What matters is being embedded in the client’s teams and tools, not being physically present every day.
Source: Angelo Lima, “Forward Deployed Engineer: the job AI borrowed from Palantir”.