If you've tried to hire a Forward Deployed Engineer (FDE) this year and found the process slower and more expensive than expected, the data says it isn't just you.
Job postings for forward-deployed engineers grew by more than 800% between January and September 2025, and 1,165% year-over-year according to Live Data Technologies. Indeed data shared with Business Insider showed a similar 729% jump in April 2026 alone - a trend running directly counter to layoffs and hiring slowdowns elsewhere in software. LinkedIn's January 2026 labour-market report found FDE roles grew 42-fold between 2023 and 2025, well ahead of the 13-fold growth in AI engineer postings over the same window.
This isn't a frontier-lab curiosity anymore. Large enterprises from Amazon to EY and Salesforce are now investing heavily in forward-deployed engineering, and OpenAI has built it into a distinct organisational function, backing it with a billion-dollar investment aimed at embedding thousands of engineers directly inside customer environments. Microsoft has gone further still, standing up a new operating division reportedly backed by 6,000 engineering specialists and $2.5 billion in investment.
Why this is a structural shift, not a fad
The driver isn't hype, it's a capability gap. The underlying problem is a widening gap between what AI models can technically do and how far enterprises have actually managed to adopt them in production. Off-the-shelf AI tooling rarely drops cleanly into a large organisation's existing systems, data governance, and workflows. Someone has to sit inside the client's environment, understand its specific constraints, and build the integration layer by hand. That's the job an FDE does; writing production code, but not for their own company's product.
The supply side hasn't kept up. FDE listings grew more than 800% while the candidate pool grew only around 50%, according to Financial Times and Live Data Technologies figures cited in a 2026 hiring guide - a gap wide enough that, as one services director put it, the problem isn't your recruiter, it's the market.
It doesn't stop at the FDE title
What matters for hiring teams is that this demand curve isn't confined to one job title. A CIEL HR study found Forward Deployed Engineer hiring up 130% over the past year, explicitly grouping the role alongside the wider set of professionals combining software engineering, AI expertise, enterprise integration and customer engagement. Live postings tracked across Palantir, Databricks, OpenAI and Anthropic show the FDE label itself splintering into specialised variants, infrastructure, reliability, security, and enablement, as companies work out which flavour of embedded AI talent they actually need.
In practice, that means the same hiring pressure is showing up under adjacent titles: applied AI engineer, deployment engineer, AI solutions architect, agentic AI engineer, field engineer. Reports tracking this space explicitly include employees in these equivalent roles whenever their day-to-day responsibilities resemble forward deployed engineering, precisely because titles are lagging the reality of what's being hired for. If your job description still says "AI Engineer" in the generic sense, you may be competing for talent against a much larger pool of postings than you realise - and losing candidates to companies who've named the role more precisely.
The skill profile hiring teams are actually screening for
The profile in demand is T-shaped: strong production engineering fundamentals (Python essential, TypeScript/JavaScript and SQL close behind) paired with hands-on experience in RAG pipelines, vector databases, and agent orchestration frameworks, plus a track record of shipping something a real customer actually used. Deep domain knowledge compounds this further; specialists who've solved the same integration problem repeatedly for one sector, such as financial services or healthcare, are described as "insanely valuable" by recruiters working this market right now. It's a narrow, specific profile, and it's exactly why generic AI-engineer sourcing is coming up short.
What this means for your hiring plan
Three things worth taking into your next headcount conversation:
- Don't assume "AI Engineer" on the job spec is enough. If your requirement is really about embedding someone inside a live production environment to close the gap between an AI capability and your actual workflows, say so - and expect to screen for the FDE/applied-AI/solutions-architect cluster, not a single title.
- Permanent hiring alone won't close this gap fast enough. With demand outpacing supply by this margin, flexible contract and interim engagement is often the only realistic way to get this capability in place within a normal project timeline - which is exactly why we're seeing sustained enterprise demand for senior contract AI engineers across sectors that aren't AI-native at all (energy, financial services, public sector).
- Compensation benchmarks have moved fast. Median total compensation at frontier labs for mid-level FDEs now starts around $385K, and broader market postings put salary medians around $173,816 with most roles including equity. Budgeting against last year's numbers will cost you candidates.
The companies moving fastest on this aren't necessarily the ones with the biggest AI budgets - they're the ones willing to look past the exact job title and build a sourcing strategy around the actual skill cluster the market is now demanding.