Forward Deployed Engineers: The AI Talent Race Moves to Deployment

It would be easy to file the forward deployed engineer under Silicon Valley job-title invention. The model was pioneered at Palantir and spread rapidly across frontier AI and enterprise technology. Our latest research says it is now spreading much further than that.

Forward Deployed Engineering is not a technology-industry trend. It is becoming an enterprise AI deployment model across sectors. As organizations move AI from experimentation to production, the hard problem is increasingly the same whether you are in retail, pharma, industrials, financial services or healthcare: how do you embed AI into real workflows, proprietary data, systems, controls and operating environments? That is precisely the gap FDE talent is emerging to solve.

For talent leaders this matters for a simple reason. The next AI talent competition may not simply be about who can build AI. It may be about who can successfully deploy it into the business. The people who can do that are scarce, and the largest pool they come from is one you already employ.

What a forward deployed engineer actually does

A forward deployed engineer sits between an AI product and the environment it has to work in. Our report maps the role across four stages: technical discovery and scoping before implementation; designing, building and integrating AI into the customer's environment; production rollout and technical enablement; and driving adoption while feeding what is learned back into the product roadmap.

Two numbers describe the job better than any title. FDEs spend 47% of their workweek embedded with customers. And only around 6% primarily build greenfield products, while 64% focus on deploying and customizing AI products that already exist. Most of the job is deployment, not invention.

That shows up in the tools. As FDEs move AI into production, adoption of evaluation and observability tooling among them grew from 52% to 78% between 2025 and 2026. The report reads this as AI evaluation replacing traditional QA: a shift from testing software features to continuously validating how an AI system performs.

Demand for deployment talent is outrunning demand for builders

Between Q1 2024 and Q1 2026, demand for Forward Deployed Engineers grew 380%, the steepest rise of any role in our comparison. Agentic AI Engineer demand rose 340%, Generative AI Engineer 280% and Applied AI Engineer 240%. Further down, ML Engineer demand grew 68%, Data Scientist 22% and Deep Learning Engineer 5%, while demand for Solutions Engineers fell 14% and NLP Engineers fell 22%.

Read that list as a talent leader and the pattern is hard to miss. The growth is concentrated in roles that put AI to work in agents, products and workflows, with the deployment role at the top. Narrower specialist roles have slowed: Deep Learning Engineer demand is close to flat, and NLP Engineer demand is down. The report's reading is that competitive advantage is shifting from model development to enterprise-scale implementation.

Adoption tells the same story. In Pave's Data Lab report covering 10,083 companies, the share with an FDE role rose from 0.34% in 2023 to 1.24% in 2026, a 3.6X increase. More than 137 companies actively hired forward deployed engineers in 2026, and nearly 8 in 10 FDE leaders expect their teams to double in size by 2027.

Why this is not only a technology-sector story

The most important finding for leaders outside tech is where the demand is heading. Around 65% of active FDE hiring now comes from enterprise software companies rather than frontier AI labs, and the top 10 employers account for nearly 40% of it. Those leaders are shaping the market ahead of broader adoption, and the report already shows the capability moving into industrial and enterprise environments.

The skills shift with the sector. In the US, tech peers such as Microsoft, Amazon and Google emphasize agent orchestration, RAG and MCP integration in their forward-deployed talent, while GM and Tesla focus on edge AI and embedded systems. In India, TCS, IBM and Accenture emphasize agentic workflows, functional solution design and fit-gap analysis, while Intelizign and Chevron focus on digital twins, industrial IoT and predictive maintenance. In the UK, Arm and Jaguar Land Rover focus on on-device inference, embedded deployment and automotive CI/CD.

The specific skills will differ. Industrial companies may emphasize edge AI, digital twins and IoT, while other sectors will require different domain capabilities. The underlying model is consistent: deep domain knowledge, engineering, deployment and business problem solving, combined in one role.

The supply gap, and where the talent will come from

Draup estimates roughly 68K relevant forward-deployed software professionals globally, about 24.2K of them in the US, 10.4K in India and 2.8K in the UK. Demand runs ahead of supply in every major metro we analyzed. The San Francisco Bay Area has about 2.0K relevant professionals against demand for 3.7K. Bengaluru has 1.7K against 2.3K. London has 1.6K against 2.3K. Across the Bay Area, New York City and Bengaluru, the demand-supply gap ranges from roughly 27% to 47%.

Large buyers are already moving at scale. TCS is planning to hire up to 8,900 forward deployed engineers. OpenAI committed $4B to a standalone Deployment Company and acquired Tomoro's roughly 150 forward-deployed engineers in May 2026. The report also puts the premium for moving from Solutions Engineer to FDE at 35%.

The more useful finding for a CHRO is where FDEs come from. Tracking employment histories, the largest single source of people moving into FDE roles is the software engineer, with 4,480 transitions. Solutions and technical architects follow at 840, data scientists at 785, AI engineers at 470 and machine learning engineers at 425. Software engineers account for more transitions than the other six feeder roles in the report combined.

The skills data points the same way. Of the four forward-deployed archetypes we analyzed, the Product Deployed Software Engineer sits closest to traditional software engineering, which creates the largest capability gap to FDE and makes it the most natural cohort for targeted upskilling. Notably, 4 of the 20 skills newly entering FDE role profiles are generative AI or agentic engineering tools. The rest formalize customer discovery, deployment governance and commercial-ownership disciplines. That suggests the reskilling path for a software engineer is as much about discovery and stakeholder judgment as it is about agentic tooling.

Buy now, build next

The report's recommendation is a sequence rather than a single move: acquire external talent in the short term, then build an embedded, in-house capability over the medium term.

• In the next one to two years, establish a competency framework covering AI depth, customer engagement and domain expertise, and use it to guide hiring. Use competitive, equity-led compensation, bring in external deployment partners selectively to bridge near-term capacity gaps, and set up executive-owned FDE pods focused on your highest-value AI use cases.
• Over three to five years, shift from a buy to a build strategy by upskilling software talent into FDE roles. Introduce specialist tracks as they emerge across infrastructure, agents, evaluation and industry domains, and reskill senior FDEs toward architecture, governance and AI value realization.

Plan the exits as well as the entries. The report sees FDEs moving into AI Deployment Lead, AI Transformation Lead and AI Solutions Architect roles, and in industrial settings into Industrial IoT Architect and Industrial/Physical AI Engineering Lead roles.

Two questions are worth taking to your next workforce planning review. Which of your current software engineers are closest to this role today? And who owns AI deployment in your business? Among companies surveyed for the report, 44% house FDEs inside Solutions Engineering and 33% distribute them across functions, while 16% run a standalone FDE team. Where you place the role is likely to shape who you can grow into it.

‍