The FDE Bubble: Silicon Valley’s Most Expensive Game Of Pretend
Silicon Valley enterprise AI ventures are securing massive funding by using human forward-deployed engineers instead of true autonomous software, creating a staffing agency model disguised as high-margin recurring revenue.
By Anurag Gurtu
Congratulations to the enterprise artificial intelligence venture that recently secured another nine-figure funding round to resolve an issue software is inherently designed to manage independently: prompting users to actually adopt the technology.
You recognize the type. No names are necessary. You have likely scanned the press release — featuring buzzwords like “autonomous,” “agentic,” and “category-defining” in the opening paragraph—only to find where the capital is truly heading a few paragraphs down: financing an army of engineers stationed inside client offices to manually force the “autonomous” software to function correctly.
Nothing screams “we’ve engineered the future of workplace productivity” quite like acquiring half a billion dollars to enlist more humans to execute the labor.
The business model, decoded
Should the investor pitch deck have left you guessing, the operational breakdown unfolds as follows:
- Market an “AI platform” to the buyer.
- Discover that the platform lacks the capability to operate autonomously.
- Deploy forward-deployed engineers: talented computer science graduates who dedicate half a year manually designing the exact workflows the marketing collateral implied were pre-built.
- Invoice the client for “software.”
- Record the transaction as Annual Recurring Revenue (ARR).
- Secure a larger funding round on the premise that “ARR metrics are rising.”
- Continue the cycle indefinitely, maintaining an expanding and costly technical staff positioned directly between the product and its marketing claims.
This structure is far from a self-sustaining business flywheel; it merely represents a payroll disguised behind a presentation deck.
Furthermore, the market has ceased hiding this reality, choosing instead to fundraise aggressively on the very premise.
OpenAI committed up to $1.5 billion toward a deployment joint venture valued at $10 billion, alongside an additional $4 billion contributed by TPG, Bain Capital, Advent, Brookfield, and Goanna. The entity, known as DeployCo, has reportedly accumulated upwards of $4 billion in total investments.
Concurrently, hiring listings for “forward deployed engineer” roles have surged between 800% and 1,000% year-over-year. Lightcast metrics published via Fortune indicated an increase exceeding 1,000% spanning January to August 2026. Financial Times figures distributed through PYMNTS registered an 800% growth in job openings between January and September 2025, while Indeed metrics tracked an expansion from 643 listings in April 2025 to 5,330 in April 2026. Overall compensation packages are surpassing $300,000 domestically. Current employer openings place Anthropic base salaries between $280,000 and $320,000, and OpenAI base pay from $185,000 to $300,000, alongside equity grants. Fortune noted select Anthropic positions achieving total compensation packages of $400,000.
Somewhere in the venture capital ecosystem, an executive is labeling this phenomenon “operating leverage” with a straight face.
This dynamic does not constitute operating leverage. It is simply a traditional staffing agency that has learned to adopt the vocabulary of “agentic” systems.
The tell
Consider posing the single inquiry that promotional decks consistently avoid: “If the embedded engineer were pulled out tomorrow, would the client still retain the value they purchased?”
If the truthful response is “well, no, but the ARR exists,” that outcome does not reflect product-market fit. It indicates people-market fit, and people-market fit operates under an unforgiving limitation: human talent can only be recruited at a finite speed, and labor can never be acquired cheaply enough to yield a true software market valuation multiple. Every subsequent client acquisition does not translate to higher operational efficiency; instead, it triggers another staffing requisition.
Consider two distinct enterprises, both boasting $20 million in ARR. One enterprise requires 10 engineers to sustain operations, whereas the opposing enterprise demands 100. Both marketing presentations label the businesses as “AI software firms.” The underlying divergence surfaces immediately within gross margins, customer onboarding expenses, and whether those operational costs decline as the organization expands.
The bill always comes due
The core characteristic of people-market fit is that it appears prosperous until critical inquiries emerge: What is the fully loaded cost required to onboard the subsequent client? Does that expenditure metric decrease as the firm scales, or does it merely bloat the corporate hierarchy? How are gross margins impacted if initial growth decelerates and the headcount can no longer be concealed behind rapid expansion?
Firms financing their forward-deployed engineering teams through fresh nine-digit financing rounds are wagering that these inquiries will remain unasked until after the transaction closes and the celebrations conclude. That approach functions adequately during market hype cycles, yet proves disastrous within a financial environment beginning to scrutinize the contractual fine print.
Eventually, market participants review the fine print. A corporate CFO constructs a true cost-per-deployment financial model. A chief information officer evaluates time-to-value benchmarks across three separate vendors rather than relying on a single demonstration. An investor finally conducts the rigorous analysis required to separate software-driven ARR from human-operated ARR on corporate financial statements.
The variance that buyers and financial backers can evaluate ultimately rests upon quantitative metrics. Genuine software ARR ought to incur lower delivery costs as volume scales. Human-reliant ARR yields no such efficiency. OpenAI notably guaranteed the private equity stakeholders of its deployment vehicle a fixed annual return of 17.5% across a five-year horizon, representing a financial structure fundamentally distinct from standard venture growth models.¹
Consequently, your artificial intelligence platform may not function autonomously. However, your underlying business model might be—automatically transferring the operational expenses typical of a professional services firm onto a software valuation multiple for the duration of the market cycle.
Enjoy the financing round. Hopefully, the expanded organizational chart proves worthwhile.
Frequently Asked Questions
01What is a forward-deployed engineer (FDE)?
A forward-deployed engineer is a technical professional hired by a software or AI company to work directly inside a customer’s office, manually building workflows and ensuring the product functions correctly.
02Why are FDEs controversial in the AI industry?
While companies market their products as autonomous software, the heavy reliance on human engineers to make the technology work resembles a staffing agency model rather than scalable software, impacting profit margins and true product-market fit.
03How fast are FDE job postings growing?
Data from sources like Lightcast, Financial Times, and Indeed show year-over-year increases in FDE job postings ranging from 800% to over 1,000%, with total compensation packages frequently exceeding $300,000 in the U.S.
Anurag Gurtu, the co-founder and CEO of Airrived, is an entrepreneur and product leader with more than 20 years of experience across cybersecurity and AI. He has built three generative AI companies, including one that reached unicorn status. Gurtu is known for turning advanced AI into scalable, real-world platforms, and previously contributed to successful exits, including acquisitions by Splunk, Tripwire and Sun Microsystems, and a public company, FireEye. He holds a master’s degree from the University of Southern California and CISSP certification.
Illustration: Dom Guzman
