Why Y Combinator and Aaron Epstein Are Betting on AI-Native Agencies

Q: Why is Aaron Epstein focused on AI-native agencies?

Aaron Epstein's background as co-founder of Creative Market gave him firsthand experience with how creative and professional services are delivered. As a YC Group Partner, his thesis is that AI has reached the capability level needed to automate the execution layer of professional services, enabling companies that combine the scalability of software with the revenue characteristics of services.

Q: Are AI-native agencies actually venture-scalable?

Yes, increasingly so. AI-native agencies can grow revenue 5-10x without proportionally growing headcount while maintaining 65-80% gross margins. This looks more like a SaaS business than a traditional services firm. The total addressable market in professional services is over $700B in the US alone, creating conditions for venture-scale returns.

Y Combinator, the world's most influential startup accelerator and the launchpad behind companies like Airbnb, Stripe, and Dropbox, has placed a massive bet on AI-native agencies. Through their Request for Startups (RFS) and the vocal advocacy of Group Partner Aaron Epstein, YC is signaling that the $700B+ professional services market is about to be fundamentally disrupted by AI-native delivery models. This is not a fringe thesis or a speculative bet on far-future technology. It is happening now, and the smartest money in Silicon Valley is positioning accordingly.

If you are unfamiliar with the concept, read our full breakdown of what an AI-native agency actually is. In this article, we go deeper into why YC specifically is betting on this model, what Aaron Epstein's thesis looks like, and what it all means for founders, operators, and the broader professional services industry.

Y Combinator's Request for Startups: AI-Powered Services

Y Combinator's Request for Startups is not a casual list of interesting ideas. It is a carefully curated signal to the global founder community about where YC believes the largest, most transformative opportunities exist. Historically, the RFS has identified categories years before they became mainstream — including fintech infrastructure, developer tools, and biotech platforms. When YC adds a category to the RFS, it is telling thousands of the world's most ambitious founders: build here.

The inclusion of AI-powered professional services alongside areas like defense technology, nuclear energy, and spatial computing signals how significant YC believes this shift to be. This is not about building another SaaS dashboard for agencies. The specific language YC uses is deliberate: they are looking for companies that use AI to deliver professional services at software-like margins. The distinction matters enormously. YC is not funding tools that help traditional agencies work slightly faster. They are funding entirely new companies that replace the traditional agency model wholesale, delivering the same or better outcomes with fundamentally different cost structures and operational architectures.

This is a departure from the prevailing VC wisdom of the last two decades, which held that services businesses were uninvestable. The fact that YC has broken from this orthodoxy tells us something profound about how AI is reshaping the economics of service delivery.

Who Is Aaron Epstein?

Aaron Epstein is a Y Combinator Group Partner who has become one of the most vocal and influential proponents of the AI-native agency thesis. His background makes him uniquely positioned to champion this category.

Co-founder of Creative Market, a marketplace for design assets that was acquired by Autodesk. This experience gave Epstein deep insight into how creative and professional services are bought, sold, and delivered at scale.
Y Combinator Group Partner, where he mentors and invests in early-stage startups across multiple batches. Group Partners are the most hands-on investors at YC, working directly with founders during the program.
Active advocate for AI-native agencies, publicly sharing his thesis through talks, interviews, and social media. He has been consistently encouraging founders to build AI-native services companies rather than pure SaaS tools.
Bridges creative services and technology. Unlike many VCs who come from pure software backgrounds, Epstein understands both the creative process and the business mechanics of services delivery. This dual perspective is why his thesis is so specific and actionable.

Epstein's core argument is straightforward: the professional services industry is one of the largest sectors of the global economy, yet it has been virtually untouched by the kind of software-driven disruption that has transformed retail, media, and finance. AI changes that equation. For the first time, it is possible to deliver knowledge work outputs — strategy documents, marketing campaigns, legal analysis, financial models — with a cost structure that looks more like software than labor.

The “Software Margins” Thesis

At the heart of the YC AI-native agency thesis is a simple but powerful economic argument about margins. Understanding this argument is essential to understanding why venture capital is suddenly interested in what was previously considered an uninvestable category.

Traditional professional services operate at 20–35% net margins. Revenue scales linearly with headcount because every new client requires proportionally more human labor. A marketing agency that wants to double its revenue must roughly double its team. Payroll, benefits, office space, and management overhead all grow in lockstep.
Software companies operate at 70–90% gross margins. The marginal cost of serving an additional customer is negligible because compute costs scale sub-linearly. This is what makes software businesses so attractive to investors.
AI-native agencies sit between these two models but trend closer to software: 65–80% gross margins are achievable and are being demonstrated by early movers. The human team handles strategy, quality control, and client relationships while AI handles the bulk of execution and production.

The key insight is that AI breaks the linear relationship between headcount and revenue. In a traditional agency, if you have 10 clients and want to take on an 11th, you likely need to hire another team member. In an AI-native agency, the 11th client adds minimal marginal cost because the AI systems that do the execution work can handle additional volume without proportional increases in compute spend. This is the economic unlock that makes VCs pay attention.

The best AI-native agencies will look like 10-person companies generating the revenue of 100-person agencies, with the margins of a SaaS business. That is a venture-scale opportunity in a trillion-dollar market.

For a detailed comparison of how these margin structures play out in practice, see our analysis of AI-native vs. traditional agencies.

Why VCs Historically Avoided Services Businesses

To understand why the YC bet on AI-native agencies is so noteworthy, you need to understand why venture capital has traditionally avoided services businesses entirely. For decades, the conventional wisdom in Silicon Valley has been clear: do not invest in services.

Services businesses do not scale. Revenue growth requires proportional headcount growth, which means capital efficiency is low.
Margins are structurally low. When your primary cost is human labor, there is a hard floor on how cheap you can deliver. You cannot compress salaries below market rate and retain talent.
Venture-scale returns are difficult. A business with 25% margins growing at 30% per year does not produce the 10–100x returns that VC fund economics require.
Moats are shallow. When your competitive advantage is talented people, those people can leave, start their own firms, or be recruited by competitors.
Key-person risk is high. Losing a star employee can mean losing their entire client book overnight.

AI-native agencies break every one of these constraints:

Non-linear scaling. AI systems handle execution, so revenue can grow 5–10x without proportional team growth.
Software-like margins. With AI handling most of the production work, margins expand from the 20–35% range to 65–80%.
AI-powered moats. Proprietary models fine-tuned on client data, custom workflows, and accumulated training data create defensible advantages that compound over time.
Data flywheels. Every project improves the AI's performance, creating a virtuous cycle where more clients lead to better outputs, which attract more clients.
Reduced key-person risk. Institutional knowledge lives in the AI systems and workflows, not in individual employees' heads.

This is why YC, a firm that has spent 20 years funding software companies, is now actively seeking out services businesses. The category has fundamentally changed.

YC-Backed AI-Native Agency Examples

While we will not speculate about specific private companies in YC's portfolio, the types of AI-native agencies that YC and similar accelerators are funding reveal the breadth of this opportunity. These are not hypothetical categories — real companies are being built and funded in each of these verticals right now.

AI-native marketing agencies that produce content at scale — blog posts, social media campaigns, ad creative, email sequences, and landing pages — using AI for generation and human strategists for direction and quality assurance. Some are reporting 10x the output per employee compared to traditional agencies.
AI-native sales agencies that automate outbound prospecting, lead qualification, personalized outreach sequences, and follow-up cadences. These firms use AI to research prospects, craft personalized messages, and manage multi-touch campaigns that would require dozens of SDRs at a traditional firm.
AI-native legal services that handle document review, contract analysis, due diligence, and regulatory compliance work. Tasks that previously required teams of junior associates billing hundreds of hours can now be completed in a fraction of the time.
AI-native accounting and bookkeeping firms that automate transaction categorization, reconciliation, tax preparation, and financial reporting. The combination of AI processing and human CPA oversight produces faster, more accurate results at lower cost.
AI-native recruiting agencies that use AI to source candidates, screen resumes, conduct initial assessments, and manage candidate communications. Human recruiters focus on relationship building and final-stage evaluation.

For a comprehensive breakdown of which verticals are most ripe for AI-native disruption, explore our guide to AI-native agency verticals.

What This Signals for the Industry

When Y Combinator bets on a category, the effects ripple far beyond the companies in their batch. YC's endorsement creates a gravitational pull that attracts talent, capital, and attention from across the technology ecosystem. Here is what the YC bet on AI-native agencies signals for the broader industry.

Talent will follow. The best engineers, designers, and operators in Silicon Valley pay close attention to YC's signals. When YC says AI-native agencies are a priority, ambitious builders start exploring the space. This talent influx accelerates the category far beyond what any single company could accomplish.

Capital will follow. Downstream investors — Series A and B firms — use YC's RFS as a signal for where to develop investment theses. Expect to see dedicated AI-native services funds and a growing number of term sheets for companies in this category over the next 12–24 months.

The professional services industry is massive. At $700B+ in the United States alone and several trillion globally, even capturing a small percentage of this market represents a multi-billion-dollar opportunity. Traditional agencies, consultancies, and professional services firms that do not adapt will face existential pressure from AI-native competitors that can deliver comparable quality at a fraction of the cost and turnaround time.

The window is closing. The best time to build an AI-native agency was six months ago. The second-best time is now. First-movers in each vertical are accumulating data advantages, client relationships, and operational expertise that will be extremely difficult to replicate once the market begins to consolidate. Early entrants who build proprietary AI workflows and accumulate training data from real client work will have compounding advantages that late entrants cannot easily overcome.

The Broader Market Context

The YC bet on AI-native agencies does not exist in isolation. Several macro trends are converging to make this the ideal moment for AI-native services companies to emerge.

Frontier AI models have crossed critical capability thresholds. Models like GPT-4, Claude, and Gemini can now produce professional-grade writing, analysis, code, and creative work. Two years ago, AI-generated content was a novelty. Today, it is production-ready for a growing number of professional use cases. The gap between AI output and human output continues to narrow, and for many routine tasks, AI output is already indistinguishable from or superior to human work.

The cost of AI inference is dropping exponentially. Inference costs have been falling roughly 10x per year, driven by hardware improvements, model distillation, and infrastructure optimization. This means that tasks which were cost-prohibitive to automate 18 months ago are now economically viable, and tasks that are borderline today will be clearly profitable within a year.

More tasks are crossing the “good enough” threshold. Every quarter, additional categories of professional work become automatable at a quality level that clients find acceptable. This expanding frontier of automation means that AI-native agencies can offer an increasingly comprehensive service portfolio over time, without proportionally increasing their human workforce.

Enterprise buyers are increasingly open to AI-delivered services. Two years ago, suggesting that a Fortune 500 company use an AI-native agency would have been met with skepticism. Today, procurement teams actively seek out AI-powered service providers because they understand the speed and cost advantages. The cultural resistance is evaporating faster than most industry observers predicted.

Remote work normalized outcome-based relationships. The shift to remote work during and after 2020 fundamentally changed how companies evaluate service providers. Clients care less about how the work gets done and more about the quality, speed, and cost of the output. This shift in buyer psychology creates a natural opening for AI-native agencies that deliver superior outcomes regardless of how many (or how few) humans are involved.

Criticisms and Counter-Arguments

No thesis is without its critics, and the AI-native agency model faces legitimate questions that deserve honest treatment.

Quality Concerns

The criticism: AI cannot match human expertise in nuanced, high-stakes work. Strategy, creative direction, and complex problem-solving require human judgment that AI lacks.

The reality: This is partially true and partially missing the point. The best AI-native agencies do not replace human judgment entirely. They use AI for execution and production while retaining human strategists for the highest-value work. The result is often higher quality than traditional agencies because the human experts spend 100% of their time on strategy and quality control rather than splitting their attention between thinking and doing. Additionally, AI quality continues to improve rapidly. Work that required heavy human editing a year ago now needs only light review.

Relationship-Driven Sales

The criticism: Professional services are sold on relationships and trust. Clients hire people, not algorithms.

The reality: AI-native agencies still have human leaders, account managers, and strategists who build and maintain client relationships. The AI handles the back-end execution. From the client's perspective, they still work with a dedicated human team. The difference is that the team is dramatically more productive because AI handles the heavy lifting. In practice, clients care about results, and AI-native agencies often deliver faster turnaround and more consistent quality, which strengthens rather than weakens client relationships.

Regulatory Risk

The criticism: Some industries have regulations that restrict or complicate AI-delivered services, particularly in legal, financial, and healthcare contexts.

The reality: This is a legitimate concern in specific verticals. However, most regulations govern the quality and accountability of the output, not the tools used to produce it. A CPA-reviewed tax return prepared with AI assistance meets the same professional standards as one prepared manually. Smart AI-native agencies design their workflows to maintain human accountability at every required checkpoint while using AI to accelerate the work between those checkpoints.

Race to the Bottom

The criticism: If every agency has access to the same AI models, margins will compress as competition drives prices down.

The reality: Access to foundation models is indeed commoditized. But the competitive advantage of an AI-native agency does not come from the base model. It comes from proprietary workflows, fine-tuned models trained on domain-specific data, accumulated client insights, and operational expertise in combining AI and human capabilities effectively. Two agencies using the same base model can produce dramatically different results depending on their prompting strategies, quality assurance processes, and domain knowledge. First-movers who build these proprietary layers will maintain defensible advantages even as the underlying models become more accessible.

Frequently Asked Questions

What is Y Combinator's Request for Startups?

Y Combinator's Request for Startups (RFS) is a public document that outlines the specific areas and categories where YC is actively looking to fund new companies. It represents YC's collective view on where the largest opportunities exist for startups. The RFS is updated periodically and serves as a signal to the global founder community about where YC believes transformative companies can be built. Being included in the RFS does not guarantee funding, but it means YC partners will be especially receptive to pitches in that category.

Why is Aaron Epstein focused on AI-native agencies?

Aaron Epstein's background as the co-founder of Creative Market gave him firsthand experience with how creative and professional services are delivered and consumed. As a YC Group Partner, he sees hundreds of startups each batch and has identified AI-native agencies as one of the most compelling opportunities in the current landscape. His thesis is rooted in the observation that AI has finally reached the capability level needed to automate the execution layer of professional services, enabling a new category of company that combines the scalability of software with the revenue characteristics of services.

Can I apply to YC with an AI-native agency idea?

Yes. YC has explicitly included AI-powered professional services in their Request for Startups, which means they are actively seeking founders building in this space. The strongest applications will demonstrate early traction (revenue, clients, or measurable output improvements), a clear understanding of the specific vertical being targeted, and a credible plan for how AI creates structural advantages over traditional delivery models. Like all YC applications, having a working product and real users significantly increases your chances.

Are AI-native agencies actually venture-scalable?

This is the central question, and the answer is increasingly yes. The venture scalability of AI-native agencies comes from their margin structure. A company that can grow revenue 5–10x without proportionally growing headcount, while maintaining 65–80% gross margins, has economics that look more like a SaaS business than a traditional services firm. The total addressable market in professional services is massive (over $700B in the US alone), and the combination of high margins, non-linear scaling, and a huge TAM creates the conditions for venture-scale returns.

What makes an AI-native agency different from an AI SaaS tool?

An AI SaaS tool gives customers software that they operate themselves. An AI-native agency delivers complete outcomes and finished work products. The client of an AI SaaS tool needs to learn the software, develop prompting skills, and integrate the tool into their existing workflows. The client of an AI-native agency simply defines what they need and receives the finished result. This distinction matters because most businesses do not want to become experts in AI tooling. They want their marketing campaigns created, their books balanced, and their contracts reviewed. For a deeper exploration of this distinction, see our comprehensive guide to AI-native agencies.

How big is the market opportunity for AI-native agencies?

The global professional services market exceeds $6 trillion, with the US market alone representing over $700 billion. This includes marketing, legal, accounting, consulting, recruiting, and dozens of other verticals. Even modest AI-native penetration of these markets represents hundreds of billions in addressable revenue. The opportunity is not just in replacing existing services but in expanding the market by making professional-grade services affordable for small and mid-size businesses that could never afford traditional agency pricing. For a breakdown of the most promising verticals, explore our vertical-by-vertical analysis.

Y Combinator's bet on AI-native agencies is not speculative. It is a calculated response to a structural shift in the economics of professional services. Aaron Epstein and the broader YC partnership see what is becoming increasingly obvious: AI has reached the capability threshold needed to deliver professional-grade work, the cost of that capability is dropping exponentially, and the companies that build AI-native delivery models now will capture an outsized share of a trillion-dollar market. For founders considering this space, the signal from YC could not be clearer. The time to build is now.