The Forward Deployed Engineer: 2026's Hottest Job Title 

When we published my 2025 predictions, we shared the rise of the Forward Deployed Engineer (FDE) as one of the trends to watch.

That prediction proved correct, but we underestimated it.

Job postings for the role rose by more than 800% between January and September 2025 (Indeed / Financial Times). Salesforce has publicly committed to hiring 1,000 FDEs. OpenAI, Anthropic, Palantir, Databricks, Cohere, Ramp, Rippling and Intercom are all building dedicated FDE functions. In April 2026, EY launched a UK & Ireland FDE practice which is the first major consultancy to formally adopt the model.

The interesting theme here is not the job title. It is the enterprise software industry quietly admitting that the old delivery model doesn’t work anymore.

What an FDE Actually Is (and Isn’t)

The role was pioneered at Palantir in the early 2010s, where they were called “Deltas.” Until 2016, Palantir had more Deltas than software engineers. The premise was that customers (governments, banks, airlines) didn’t need more product features. They needed engineers who could make the product work inside their fragmented, legacy, regulated, politically complex environments.

A Forward Deployed Engineer is a customer-embedded engineer who writes production code inside the client’s environment to make a complex product actually work in the real world.

That’s the definition. The distinction matters because the role is often confused with adjacent ones:

• Solutions Engineers sell the vision. FDEs make it real. SEs run pre-sales demos. FDEs ship the integration after the deal closes.
• Customer Success Engineers guide. FDEs build. CSEs work within what the product supports today. FDEs extend it to fit the customer’s reality.
• Consultants advise. FDEs own outcomes. Consultants leave a deck. FDEs leave a system running in production.

Palantir’s own job descriptions describe FDE responsibilities as “looking similar to those of a startup CTO.” That framing is deliberate as the role is built for high-ownership operators who thrive in ambiguity, not implementation-checklist managers.

Why the Role Has Exploded Now

The role is not new but there are three factors driving the numbers we shared:

1. AI productisation has outpaced AI deployment. Foundation models are commodifying. Enterprises have spent billions on AI in 2024 and 2025 but most pilots have not made it to production. The bottleneck isn’t the model in most cases, it’s the messy reality of enterprise SSO, legacy ETL, regulated data, and security review boards.
2. Buyers want outcomes, not licences. The 2024–25 era of “AI FOMO” is over. CFOs are now demanding measurable ROI on every AI line item. That shifts the buying conversation from features to delivered value. To get this value it requires someone on the ground who can shape the product to the desired outcomes of the business.
3. Platforms are commoditising, deployment is the moat. When everyone has access to the same models and the same SaaS layer, competitive advantage moves to the speed and quality of implementation. Whoever deploys fastest wins the renewal.

This is exactly the macro shift we have been writing about under the 40/40/20 workforce model and the Chief Work Officer concept which is the boundary between vendor and customer is dissolving, and the unit of value is becoming the embedded, outcome-accountable specialist.

The Skills That Actually Matter

As a skills platform we see specs that list 30 skills. Using the SFIA 9 skills framework you can actually break it down into:

We would expect a strong FDE to be operating at SFIA Level 5 (Ensure, advise) or Level 6 (Initiate, influence) across most of the technical skills, and at Level 4–5 across the commercial and behavioural skills. That blend is what makes the role rare.

Technical skills (SFIA 9):

• Programming/software development (PROG) — Level 5/6. One native language they can ship production code in blindfolded — typically Python or TypeScript in 2026 — and working fluency in two or three more (Java, Go, Rust, C#, sometimes Scala in banking). At Level 5 they are designing, coding and refactoring complex software components and setting standards for others.
• Software design (SWDN) and Solution architecture (ARCH) — Level 5. Full-stack engineering across APIs, databases, backend services and enough frontend to demo, with the ability to architect end-to-end inside a customer’s existing landscape.
• Systems integration and build (SINT) — Level 5. The hard part of the job: integrating a new platform with the customer’s legacy ETL, identity, data warehouse and security stack.
• Infrastructure design (IFDN) and Deployment (DEPL) — Level 4/5. Cloud platform depth across AWS, Azure or GCP, including IAM, networking and data services. Comfortable owning a production deployment from sandbox to GA.
• Machine learning (MLNG) and Data engineering (DENG) — Level 4/5. Practical AI/ML literacy: LLM orchestration, RAG, evals, agent frameworks (LangChain, LlamaIndex), vector stores. Real-world data engineering — modelling, pipelines, integration with Snowflake, Databricks or Salesforce Data Cloud.
• Problem management (PBMG) and Application support (ASUP) — Level 5. Ability to read an unfamiliar codebase, diagnose a production issue, and ship a patch against it inside a week.

Commercial and behavioural skills (SFIA 9):

• Requirements definition and management (REQM) — Level 5. Translating ambiguous business problems into technical specs without 20 discovery meetings. This is the single skill that most distinguishes a good FDE from a strong product engineer.
• Stakeholder relationship management (RLMT) — Level 5. Reading a room of customer engineers (who often feel threatened by a vendor that deep in their stack) and customer VPs (who want unrealistic timelines) simultaneously, and managing both relationships through the deployment cycle.
• Consultancy (CNSL) — Level 5/6. Willingness to disagree in public — with the customer in front of their own sales team, and with their own product team in front of the customer. Holding a defensible position based on evidence.
• Specialist advice (TECH) and the SFIA generic attribute of Communication at Level 5. Executive presence and written communication that survives scrutiny by a CIO or CDAO.
• Customer service management (CSMG) combined with the SFIA generic attribute of Influence at Level 5. Travel tolerance, depending on the flavour: Enterprise Integration FDEs typically travel 20–40%; AI-native FDEs are mostly remote since 2023.

Compensation: An FDE is Not Cheap

Levels.fyi data current to early 2026 puts average total compensation for an FDE in the US at around $238,000, with the typical range $205,000–$486,000 and Staff-level FDEs clearing $630,000+. Palantir, OpenAI and Anthropic sit at the top of the band. The premium exists for one reason: skill scarcity. You are paying for an engineer who can also manage a multi-million-dollar relationship.

In the UK the bands are narrower but the trajectory is identical. Glassdoor data for “Forward Deployed Software Engineer” in London (April 2026) shows average total compensation of £138,000, with the typical range £108,000–£186,000 and top earners reporting £253,000+. Levels.fyi puts the UK median at £121,000. Palantir’s London FDE function — the largest in Europe — sits in the £155,000–£195,000 base band at senior levels, with significant equity on top. Outside London the market is roughly 25–35% below those numbers, though remote-first AI labs are flattening that gap quickly.

EY’s launch will accelerate the pricing reset across UK consultancies, because they now have to compete with AI labs and Palantir for the same shortlist. Expect Big Four FDE base salaries to clear £130,000 by end-2026, with Tier 2 SIs forced to follow within 12 months.

Who Is Hiring (and What Flavour They Want)

Three distinct hiring patterns have emerged:

AI labs and AI-native scaleups — OpenAI, Anthropic, Cohere, Databricks, Salesforce (Agentforce), Ramp, Rippling, Intercom, Adobe (Firefly). These FDEs are 70% remote, contribute back to the core product, and operate closer to the founder/CTO archetype. Salesforce runs them in “pods” which is one deployment strategist plus two FDEs, locked onto a single customer for ~3 months.

Enterprise platform companies — Palantir is still the gold standard. Datadog, Snowflake and the data infrastructure players have all built FDE-equivalent functions. Higher travel, deeper integration work, more regulated industries.

Consultancies and SIs — EY’s April 2026 launch is the leading indicator. Expect Deloitte, Accenture, PwC, KPMG and the Tier 2 SIs to follow within 12 months. The pitch: senior engineers embedded in client delivery teams, governance and compliance baked in from day one. This is also where the on-demand model becomes structurally important — consultancies cannot scale FDE supply through traditional hiring alone.

Where to Find Them

This is the question every CIO and CTO is now asking. There are four routes, each with trade-offs:

1. Direct hire from the AI labs and Palantir alumni network. The highest quality, the highest cost, and a 6–9 month time-to-hire. The candidate pool is narrow because most FDEs are content where they are.
2. Specialist FDE talent platforms. Sites like fwddeploy.com (the dedicated FDE job board), Second Talent, WorkGenius and BorderlessMind have emerged specifically for this role, offering pre-vetted candidates in 24–48 hours.
3. On-demand and fractional FDE talent. This is where the market is moving fastest. Most enterprises do not need a permanent FDE for every use case, they need a senior, embedded engineer for the duration of a specific deployment. Due to the hiring demand, it can be a long hiring cycle so on-demand becomes a viable option. Marketplaces of vetted independent FDEs let buyers spin up capacity per-engagement, blending platform delivery with human expertise. This is the model we operate at Gigged.AI: an on-demand marketplace of 25,000+ vetted tech specialists, including a growing FDE-shaped cohort, that enterprises tap for project-bounded work without the 9-month hiring cycle.
4. Internal upskilling. The most under-used route. Every enterprise already has senior engineers with technical depth. What they lack is the commercial muscle and the customer-facing reps. Building an internal FDE function — by retraining existing platform and solutions engineers — is faster and cheaper than the external market suggests, particularly when paired with workforce intelligence tooling that maps current skills against the FDE profile.

The right answer for most organisations is on-demand and internal in parallel. Use the on-demand market to deliver the next 90 days of customer commitments while you build the internal bench for the next 24 months. Treating it as one or the other is a false choice.

Conclusion

The Forward Deployed Engineer is not a new job. 

The companies that win the next five years of enterprise AI will not be the ones with the best models. They will be the ones with the deepest deployment capability through a blend of permanent FDEs, on-demand specialists, and internal talent that has been deliberately upskilled into the role.

If you are a CIO, CTO or COO reading this and you do not yet have a defined FDE strategy — for both your own platform and the platforms you are buying — that is the gap to close in 2026.