What is Forward Deployed Engineer?

Written By Gozde Gorce

Real Customer Success + GTM insights, in your inbox!

Every Tuesday, what's actually working in Customer Success and post-sales GTM right now.

​

Tactical frameworks for breaking in or scaling up. Honest takes on hiring, AI in CS, and where the industry is heading.

​

From a CS leader who's scaled teams at hyper-growth startups in Silicon Valley and coached 100+ career-pivoters into the industry.

​

Free, in your inbox.

    I respect your inbox. Unsubscribe at any time.

    By Gozde Gorce, founder of The Success Path, and creator of How to Break Into Customer Success, with more than 10+ years in hyper-growth startups in the Bay Area, and now coaches career pivoters into the new wave of Post Sales roles in Tech.

    CSM hiring peaked in Q2 2022. It has been flat for four years. Forward Deployed Engineer roles are up 1,165% in the same window.

    TL;DR

    • The Forward Deployed Engineer (FDE) is the role that takes the seats CSM used to hold at applied AI companies.

    • FDE hiring is up 1,165% in four years. CSM hiring stayed flat. This is not a slow shift.

    • FDE work has two functions: context (mapping data, capturing undocumented workflows) and correctness (evaluating outputs, ensuring quality). Both required.

    • The title varies by domain. Applied AI labs call it FDE. Healthcare AI calls it FDE Life Sciences. Same role architecture, different label.

    • Real hiring patterns: Harvey hires former lawyers. Intercom Finn hires former consultants. Anthropic and Lovable hire from consulting and solutions architect tracks.

    • The biggest barrier for senior CS pros is language, not skill. The vocabulary is different.

    MIT NANDA's August 2025 GenAI Divide report found that 95% of enterprise AI pilots return nothing on $30 to $40 billion in spend. The FDE role is how applied AI companies are trying to be in the 5%.

    What is a Forward Deployed Engineer?

    Forward Deployed Engineer (FDE): A role that embeds directly with a customer to configure, deploy, and own outcomes from an AI product inside that customer's environment. Pioneered at Palantir in the early 2010s, now standard at applied AI companies.

    The pure FDE track writes production code, owns delivery end-to-end, and is accountable for outcomes in a way a consultant or solutions architect is not.

    But the same architecture has spawned adjacent titles that do not require production engineering: Solutions Architect, Customer Engineer, Implementation Strategist, Technical Account Manager (TAM), Success Engineer, GTM Engineer. Some applied AI companies have eliminated the CSM title entirely and replaced it with TAM.

    What is a Forward Deployed Engineer called in your industry?

    The title varies by domain. Same role architecture, different label.

    • Applied AI labs (Anthropic, OpenAI, Scale, Cursor): Forward Deployed Engineer, Customer Engineer, Solutions Architect

    • Legal AI (Harvey, Spellbook): Forward Deployed Engineer with deep legal domain expertise

    • Customer support AI (Intercom Finn, Decagon): Implementation Strategist, Customer Engineer

    • Martech and sales tech: GTM Engineer. This is the FDE pattern combined with traditional CSM ownership. Deployment-heavy, but also responsible for outcome ownership, expansion, and renewal motion. If you are coming from a sales-adjacent CS background, GTM Engineer is your most natural target.

    • Enterprise data and analytics platforms: Forward Deployed Software Engineer, Solutions Architect

    • Healthcare and life sciences AI: FDE Life Sciences (both Anthropic and OpenAI post this title), Clinical Solutions Architect, Implementation Lead

    The takeaway: if you are job-searching, do not search only "Forward Deployed Engineer." Search the domain-specific title. The GTM Engineer search alone surfaces 5 to 10x more openings than the FDE search at martech and sales tech companies.

    Why are Forward Deployed Engineer roles up 1165%?

    Applied AI companies do not sell access to a tool. They sell the work itself. That changes what the customer-facing team has to deliver.

    Two forces are reshaping CS at the same time:

    • Technical: Non-deterministic systems. The same input no longer produces the same output. You have to verify quality, not just availability.

    • Commercial: Customers pay for outcomes, not seats. There is no NRR ceiling because there is no seat cap.

    Success requires engineering two conditions: high context and verifiable correctness. Context means the AI has the right inputs. Correctness means you can prove the output is good. Both are non-negotiable.

    The FDE role is built around delivering both inside the customer's environment. Sequoia calls this the services-as-new-software shift: $6 spent on services for every $1 spent on software. AI is eating that $6.

    What does a Forward Deployed Engineer actually do?

    The work is split into two functions within the same role.

    Context (the half that maps the customer):

    • Maps data sources

    • Captures tribal knowledge that was never documented

    • Embeds in the customer's workflow

    • Translates business needs into AI configuration

    Correctness (the half that proves the work):

    • Builds evaluation frameworks (evals)

    • Engineers prompts and agents

    • Debugs model output

    • Monitors drift over time

    The team usually staffs people who lean one way based on their backgrounds. Senior FDEs lead across both. You need to speak credibly to both, even if you start strong on one.

    How Applied AI companies are staffing Forward Deployed Engineer roles

    The hiring patterns at applied AI companies tell you exactly what backgrounds win. Real examples:

    • Harvey (legal AI) hires former lawyers. In the legal domain, context is abundant (case law is documented). The bottleneck is correctness validation, which requires legal judgment. So Harvey staffs domain experts who can grade the output.

    • Intercom Finn (customer support AI) hires former consultants. In customer support, every customer's workflow is undocumented. The bottleneck is context capture, which is the consulting muscle. Correctness is easier to measure (was the ticket resolved or not).

    • Sales Intelligence companies hire what they call GTM Engineers. Same job as an FDE, different label.

    And here is what the major AI labs explicitly say in their job postings:

    So here is the real hiring map for non-engineer FDE-adjacent roles in 2026:

    • Consulting (MBB, Big 4, boutique). Strongest entry. You synthesize undocumented workflows for a living.

    • Solutions Architect or Sales Engineering. Strongest technical bridge. You already speak APIs, integrations, and technical buyers.

    • Product Management. Increasingly common at AI-native hires. You already define quality bars.

    • Customer Success at a consumption-based SaaS company. Closest direct analog. Snowflake, Datadog, Twilio, and Databricks resumes are in demand.

    • Success Engineer, Implementation Consultant, Functional Consultant. Already technical, already customer-facing. Often the fastest path.

    • Deep domain expertise (law, medicine, finance, manufacturing, compliance) on top of any of the above is the multiplier. A consultant who is also a lawyer is gold for legal AI.

    What does that mean for YOU, if you are coming from a traditional customer success background?

    In my experience, traditional CSMs fail applied AI interviews not because of a skill gap but because of a vocabulary gap. The gap lies in transitioning the work you have done as a CSM into the new era of applied AI. They use 2019 CSM playbook language: QBR, NRR dashboard, renewal cadence, health score. The hiring manager uses AI-native vocabulary: eval pass rate, agentic workflow, drift, time to first value, velocity. The conversation breaks down before the substance is evaluated.

    Career pivoters have an unfair advantage here. You are not coming in with a decade of QBR muscle memory and renewal-motion vocabulary. You can learn the AI-native terms cold and use them naturally.

    The vocabulary to know: eval, eval pass rate, prompt engineering, agentic workflow, drift, time to first value, velocity, applied AI, deployment, context window, deterministic vs non-deterministic, FDE, TAM, GTM Engineer.

    If those words are unfamiliar, you have a four-week learning curve to fix that. Most senior CSMs are not investing in it. That is your opening.

    How to break into a Forward Deployed Engineer role

    Three concrete moves (the same ones I run my How to Break Into Customer Success students through right now):

    • Read the actual job postings at Anthropic, Lovable, and OpenAI. Highlight every word you do not use today. That is your vocabulary list.

    • Identify your entry point from the hiring map. Consulting, SA, sales engineering, product, success engineering, or consumption-based CS. Update your LinkedIn headline this week.

    • Build one eval. Take any AI workflow (Claude, ChatGPT, an agent) and write a 10-row evaluation rubric for it. Publish it. That is a portfolio piece almost nobody else has.

    Bonus move: get yourself into a hackathon. Lovable, Clay, Anthropic, and OpenAI all run them. This is where 26-year-olds are getting hired directly into FDE-adjacent roles right now, and where senior CSMs are not showing up.

    Should you pivot into a Forward Deployed Engineer role?

    The math is clear. CSM hiring has been flat for four years. FDE hiring is up 1,165%. The function is being rebuilt around context and correctness, and the new title is not CSM.

    You do not have to unlearn anything. You only have to learn the language and pick your entry point.

    Frequently asked questions about Forward Deployed Engineer roles

    What is a Forward Deployed Engineer?

    A Forward Deployed Engineer (FDE) is a role that embeds with a customer to configure, deploy, and own the outcomes of an AI product within that customer's environment. The role was pioneered at Palantir in the early 2010s and is now standard at applied AI companies. The pure FDE track requires production engineering. Adjacent titles (Customer Engineer, Solutions Architect, TAM, Success Engineer) often do not.

    How is a Forward Deployed Engineer different from a CSM?

    A traditional CSM manages renewals, QBRs, and seat adoption. An FDE owns deployment, evaluation, and outcomes inside the customer's environment. CSM hiring has been flat since Q2 2022. FDE hiring is up 1,165% in the same window. At many AI-native companies, the FDE role has replaced the CSM seat entirely.

    Do I need to code to become a Forward Deployed Engineer?

    It depends on the title. Production FDE roles at companies like OpenAI require 5 to 8+ years of engineering. FDE-adjacent roles (Customer Engineer, Solutions Architect, TAM, Success Engineer) hire from consulting, sales engineering, product, and domain expert backgrounds without production coding requirements. In this era, to code, you can use tools like Codex, Claude Code, Replit, or Lovable to build an MVP from scratch without needing to write one line of code yourself.

    What background gets hired into Forward Deployed Engineer roles?

    Consulting, solutions architecture, sales engineering, product management, success engineering, and customer success at consumption-based SaaS. Domain expertise (law, medicine, finance) is a multiplier on top. Harvey hires former lawyers for legal AI. Intercom Finn hires former consultants for support AI. Anthropic hires from the consulting and solutions architect track. My two cents as someone who has been on the hiring side: as long as you are a builder, like to explore new solutions, and have domain expertise, it is all about how you convey these two in interviews, which is where most people fail.

    What is the difference between a Forward Deployed Engineer and a Solutions Architect?

    A Forward Deployed Engineer typically embeds inside a customer environment for an extended deployment, writes code, and owns outcomes end-to-end. A Solutions Architect leads technical discovery and design earlier in the sales cycle and may not stay through deployment. At many applied AI companies, the two roles overlap heavily.

    Where do I find Forward Deployed Engineer jobs to apply for?

    Search job titles like Forward Deployed Engineer, Customer Engineer, Solutions Architect, Sales Engineer, Technical Account Manager, Implementation Strategist, and GTM Engineer. Applied AI companies (Anthropic, OpenAI, Scale, Glean, Harvey, Hebbia, Intercom, Lovable, Cursor) and other AI-native startups are the highest-density employers.

    Are Forward Deployed Engineer jobs only at AI companies?

    The 1,165% growth is overwhelmingly driven by applied AI companies, but the role predates the AI wave. Palantir invented it. . The current surge is AI-native.

    Resources cited in this post

    Gozde Gorce
    Next

    Drive Engagement with Stakeholder Alignment as a CSM in 3 Easy Steps