FDE vs Solutions Engineer vs Implementation Engineer

The forward deployed engineer (FDE) title has spread quickly with AI. It is often confused with solutions engineering and implementation engineering. The roles overlap but they are not the same.

Forward deployed engineer

An FDE is a software engineer who works directly with a customer to deploy a complex product or system into that customer's environment, and who owns the outcome. They write production code, design integrations, handle the customer's data and constraints, and feed what they learn back into the core product. The role emerged at data and enterprise software companies and is now common at AI companies whose products only create value once integrated.

Distinguishing marks: builds and ships code in the customer's context; owns a deployment end to end; works across product, data and integration; usually embedded for weeks or months.

Solutions engineer

A solutions engineer supports the sales cycle: discovery, demos, proofs of concept, technical answers to RFPs. They are deeply technical but their output is mostly pre-sale. After the deal closes, delivery usually moves to another team. Some organisations blur the line by having solutions engineers run pilots.

Distinguishing marks: pre-sales focus; demos and POCs; breadth over depth; measured on pipeline and win rate.

Implementation engineer

An implementation engineer configures and integrates a product for a customer after the sale, typically following a known playbook. The work is closer to configuration and integration than to novel engineering, and it is usually time-boxed per customer.

Distinguishing marks: post-sales; repeatable playbook; configuration and integration; measured on time to go-live.

Why AI made FDEs important

AI systems rarely work out of the box in a specific business. Retrieval must be tuned to real documents, agents must call real systems, evaluation must reflect real tasks. That work needs an engineer who can change the product, not just configure it, and who sits close enough to the customer to see what is failing. That is the FDE.

What buyers should take from this

If you are buying AI deployment help, ask which of these roles you are actually getting. A provider offering "FDEs" who only configure a platform is selling implementation engineering. A provider offering engineers who will change code, own integrations and hand over evaluation sets is selling forward deployed engineering. Both can be right; they cost and deliver differently.