Cloud Intelligence™
What is a Forward Deployed Engineer?
A Forward Deployed Engineer (FDE) embeds inside a customer's environment to build, ship, and maintain technical fixes. Learn what FDEs do, how the role differs from a solutions architect, and how DoiT's FDE team works.
This page is also available in Deutsch, Español, Français, Italiano, 日本語, and Português.
About Josh Palmer
I'm Josh Palmer, Head of Content at DoiT, where I split my time across multiple business units including DoiT Cloud Intelligence, PerfectScale (Kubernetes cost optimization), and SELECT (Snowflake, Databricks, and BigQuery cost optimization). Before DoiT, I spent four and a half years at OnBoard building content for a board intelligence platform used by 6,000+ organizations, and before that, two years as Content Marketing Manager at Zylo, a SaaS management platform.
My personal pageTL;DR: A Forward Deployed Engineer (FDE) is a hands-on engineer who embeds directly inside a customer's environment to build, integrate, and maintain a technical solution, rather than handing off a design document and moving to the next account. FDEs own the full arc of a deployment: discovery, integration, debugging, and long-term support. DoiT runs one of the larger FDE practices in cloud and FinOps, pairing senior cloud architects with customer engineering teams to ship fixes for cost, reliability, and security instead of just flagging them.
Your cloud bill keeps climbing, and the dashboard your team bought last year just adds one more alert nobody has time to chase. Someone still has to read the trace, figure out why a Kubernetes node pool won't scale down, and ship the fix before the next invoice lands. That gap between "here's the problem" and "here's the fix, deployed" is exactly what the Forward Deployed Engineer role exists to close.
What does a Forward Deployed Engineer do?
A Forward Deployed Engineer works embedded with specific customers, applying deep technical skills across whatever mix of infrastructure, data, and integration work those customers need. A typical software engineer owns one product capability across many customers. An FDE flips that model: they own a small set of customers across many capabilities, staying close enough to each environment to understand its architecture, its business context, and its constraints.
That embedded posture shapes the job. An FDE's responsibilities usually span the full lifecycle of a deployment rather than a single phase of it: requirements discovery and technical scoping, system integration and implementation, deployment and rollout, ongoing debugging as real usage surfaces edge cases, and enablement so the customer's own team can sustain the work after the FDE moves to the next priority. A solutions architect typically exits after the design gets approved. A consultant typically exits after the report gets delivered. An FDE stays through production.
Where did the Forward Deployed Engineer title come from?
Palantir Technologies popularized the term, using it to describe engineers who deploy software directly inside government and enterprise environments where the domain complexity is too high to solve from a distance. As generative AI vendors and infrastructure companies took on similarly complex, high-stakes deployments, the title spread across the industry. OpenAI now hires directly for the "Forward Deployed Engineer" title, while other cloud providers use different names for comparable roles: Google Cloud calls it "customer engineer," and Amazon Web Services uses "solutions architect." The responsibilities overlap, but the depth of embedding and the length of the engagement often differ by company and by role.
How is a Forward Deployed Engineer different from a solutions architect or a consultant?
All three roles help a customer get technology working. The difference is when they show up and how long they stay.
| Role | When they're involved | What they own | When they leave |
|---|---|---|---|
| Solutions architect | Pre-sale and design phase | Architecture recommendations | After the design gets approved |
| Consultant | A fixed engagement window | A report or roadmap | After the report gets delivered |
| Forward Deployed Engineer | Full deployment lifecycle | Implementation and measurable outcomes | Stays embedded through production and iteration |
That last row is the part that trips people up. An FDE isn't a support technician working a ticket queue, and they aren't a consultant handing over a slide deck. They sit inside the account, alongside the customer's own engineers, and they stay accountable for the outcome after the initial build ships.
What skills does a Forward Deployed Engineer need?
The technical bar is high because the job spans so many disciplines at once. Most FDE roles require hands-on production experience with at least one major public cloud (AWS, Google Cloud, or Azure), comfort with infrastructure as code such as Terraform or CloudFormation, and working proficiency in a general-purpose language like Python, Go, or TypeScript. Container platforms, particularly Kubernetes, show up constantly, and observability tooling matters just as much as the infrastructure itself: an FDE needs to reason about cost, performance, and reliability trade-offs, not just implement a fix. As AI workloads move into production, GPU utilization, inference cost control, and model deployment have become a standard part of the FDE skill set rather than a specialty add-on.
The technical skills get an FDE in the room. The soft skills keep them useful once they're there. Translating a vague pain point like "our AWS bill spiked again" into a concrete, scoped technical plan takes practice, and so does explaining that plan to a finance stakeholder who doesn't care about node pools and an engineer who does. FDEs also work inside someone else's account team, coordinating with customer success managers and account managers rather than owning the customer relationship outright.
Cloud bill shouldn't be a mystery
One platform for AI and Cloud optimization.
What does a Forward Deployed Engineer team look like in practice? DoiT's FDE model
DoiT, a multicloud FinOps and cloud operations company, built one of the larger dedicated FDE practices in the space. The team started as Customer Reliability Engineering (CRE) and was renamed Forward Deployed Engineering to match how the market already understood the role: engineers embedded with customers to deliver integrated outcomes, not a support desk that closes tickets. The rebrand didn't change the underlying job. It changed the framing, because "FDE" carries a clearer signal to both customers and the engineers DoiT wanted to hire.
DoiT positions the practice directly against the alternative. As the team's own materials put it, "Not a ticket queue. Not a consulting deck." DoiT's FDEs are senior cloud architects, working across nine time zones, who pair with a customer's platform engineering and FinOps teams to ship changes rather than send alerts. The practice organizes around six disciplines: AI and LLM optimization, cost optimization, incident response, Kubernetes tuning, migration support, and reliability and performance. Individual FDEs also carry deep specialization, pulled into an engagement based on what the customer's stack actually needs.
Example FDE specialties at DoiT
- Rajan Bhave leads GenAI accelerators across EMEA, designing the Bedrock and AgentCore reference architectures that take a customer from demo to a production agent.
- Sascha Heyer works the ML and LLM side, building Vertex AI pipelines and model deployment workflows.
- Eduardo Mota also covers ML and LLM work, focused on Bedrock-based RAG systems and getting generative AI into production.
- Chimbu Chinnadurai, known internally as the Kubernetes whisperer, spends his days getting misbehaving EMEA clusters to behave and figuring out why they misbehaved in the first place.
- Kate Gawron covers the database side, modernizing Aurora, RDS, and Snowflake deployments.
Those five are just a sample. DoiT's FDE practice runs roughly 200 engineers deep, so the range above barely scratches the surface, but it's enough to show the pattern: a customer engagement can pull in the specialist it actually needs instead of a generalist guessing at an unfamiliar stack.
Attribute by DoiT is a concrete example of how that plays out. Shared infrastructure such as GPUs, Kubernetes clusters, and multi-tenant services tends to hide who's actually driving a given cost, and Attribute uses eBPF telemetry to map spend back to the team, workload, model, feature, or customer responsible for it. Getting there isn't a one-time install. DoiT's FDEs deploy the eBPF collector, validate that the telemetry lines up with the customer's actual cloud bills, design an attribution model that matches how the business defines its teams and tenants, and then help the customer's own team wire that output into its FinOps workflow over a defined 90-day onboarding. That sequence, discovery, build, validation, and ongoing enablement, is the FDE lifecycle in miniature.
When should a company bring in a Forward Deployed Engineer?
A few signals tend to show up together when a Forward Deployed Engineer is the right call. Alert fatigue is one: a team has plenty of dashboards flagging problems and no one with the bandwidth or context to fix them. Cost or reliability issues that need code-level or configuration-level changes, not another report, are another. Shared or multi-tenant infrastructure that blends many cost drivers into one bill makes the case stronger, and GenAI or LLM workloads with unpredictable, fast-moving spend make it stronger still. Cloud migrations also benefit, since context gets lost every time a project changes hands between teams, and an embedded engineer who stays through the whole move closes that gap.
Frequently asked questions
Is Forward Deployed Engineer a real job title, or just internal jargon? It's a real, external-facing job title. Palantir popularized it, and it's now used across AI infrastructure companies like OpenAI and cloud and FinOps vendors, including DoiT, to describe engineers who embed with customers rather than sell from a distance.
What's the difference between an FDE and a solutions architect? A solutions architect typically advises during the pre-sale or design phase and moves on once the architecture gets approved. An FDE stays through implementation, deployment, and ongoing iteration, owning outcomes rather than just recommendations.
Do Forward Deployed Engineers write production code? Often, yes. Depending on the company and the engagement, FDEs commonly build infrastructure-as-code modules, automation scripts, and integration code, and in some organizations they contribute directly to the vendor's own product under standard engineering review.
What skills do you need to become a Forward Deployed Engineer? Hands-on production experience with at least one major cloud provider, infrastructure-as-code fluency, a general-purpose programming language, and strong communication skills for translating business pain points into technical plans. Depth in one domain, such as Kubernetes, data platforms, or ML/GenAI, is typically required on top of that broad base.
How is an FDE different from a traditional support or help desk model? A support desk responds to tickets as they come in. A Forward Deployed Engineer works proactively inside the customer's environment, identifying and shipping fixes before they become tickets, and staying accountable for the outcome rather than closing a case.
See how DoiT's Forward Deployed Engineers work inside a live cloud environment: book a 15-minute call to map your CloudOps priorities.
