The latest trendsetter from YC: 9 startups favored by top VCs

Every year, the YC (Y Combinator) Demo Day serves as the "wind vane" for the global venture capital and startup ecosystem. Companies like Airbnb and Dropbox emerged from here, and now it has become a stage for AI startups to showcase their capabilities.

Last week, the YC Summer 2025 Demo Day came to an end, with over 160 startups making their debut. Different from previous batches, the trend this year is clear: Many companies are no longer focused on "AI-driven" products but are building AI agents or developing the infrastructure and tools required for these agents.

Silicon Rabbit and several investors focused on YC startups had conversations to find out which startups they thought were the most interesting and which ones received the highest investment attention. We hope this can be a starting point for you to gain insights and initiate deeper thinking.

01

Autumn

Business: The Stripe (payment solution) in the field of AI startups.

Why it's favored: Many AI startups use complex pricing models, usually a mix of fixed subscription fees per seat, usage-based billing, credits, and various additional fees. Managing such complex AI pricing on Stripe is a time-consuming and manual process.

Therefore, Autumn has developed an open-source infrastructure that simplifies the integration with Stripe for AI startups. The company claims that its technology has been used by hundreds of AI applications and 40 YC startups. Given Stripe's dominance in the payment field and the explosive growth of the AI market, could a billing solution designed specifically for AI become the next major fintech success story?

Dedalus Labs

Business: The Vercel (front-end development and deployment platform) in the field of building AI agents.

Why it's favored: Just as Vercel helps developers deploy and host websites, Dedalus Labs claims that its platform can automate the infrastructure required for building AI agents, reducing hours of coding work to just a few clicks. The company handles complex tasks such as automatic scaling and load balancing and says it makes deploying AI agents fast and easy.

Design Arena

Business: Crowdsourcing rankings for designs with a certain "vibe".

Why it's favored: AI can quickly generate a large number of designs, which brings a new problem: How to determine which designs are truly excellent. Design Arena solves this problem by crowdsourcing rankings for AI-generated visual works, creating a feedback loop that forces AI models to improve. Large AI labs see the value in training their models to generate better designs, and some of them are already clients of Design Arena.

Getasap Asia

Business: A technology-driven distributor for Southeast Asian retailers.

Why it's favored: Getasap Asia was founded three years ago by Raghav Arora when he was only 14 years old. Since then, this startup, which uses technology to deliver supplies to corner stores, restaurants, and large supermarkets in Southeast Asia within eight hours, has generated millions of dollars in revenue. According to its official website, Getasap Asia has completed a round of financing from top-tier capital General Catalyst, and we've heard that the company's valuation is one of the highest among this batch of YC projects.

Keystone

Business: An AI engineer that fixes software bugs in production environments.

Why it's favored: Keystone was founded by 20-year-old Pablo Hansen, who just obtained a master's degree in AI last year. The company's mission is to reduce software crashes. The company's AI finds and fixes bugs for clients like Lovable, and Hansen says they've rejected a million-dollar acquisition offer.

RealRoots

Business: An AI matching platform for women to find friends.

Why it's favored: While dating apps are everywhere, RealRoots is addressing another kind of loneliness. The company's AI matchmaker "Lisa" interviews women and then organizes social experiences to connect them with like-minded friends.

Although the AI part might be more of a "show" - conversations with Lisa may not provide RealRoots with more insights about participants than written Q&A - RealRoots might be on the right track. Its founder says that just last month, the company earned $782,000 from 9,000 paying customers.

Solva

Business: Automating insurance claims with AI.

Why it's favored: Solva's AI automates the most routine tasks for insurance claims adjusters, from filling out complex claim forms to preventing improper payouts. Just ten weeks after its launch, Solva's Annual Recurring Revenue (ARR) reached $245,000, a figure that has excited investors.

Pingo AI

Business: An AI foreign language tutor.

Why it's favored: Apps like Duolingo have made language learning easy and fun, but they usually lack a key element for fluently mastering a language: continuous conversation. Pingo solves this problem by allowing users to have oral conversations with its AI that acts as a native speaker. The company's unique approach is becoming extremely popular, and its founder claims that the business is growing by 70% monthly, with a monthly revenue of $250,000.

02

The above is a quick overview of the most eye-catching projects. However, for astute investors, the value of information goes beyond this. Looking beyond the surface, Silicon Rabbit and its expert team have distilled four core trends, which might be the best entry points to understand the current innovation trends in Silicon Valley.

1: From "Empowerment" to "Native" and "Agents"

The most obvious signal at this YC is that startups are no longer satisfied with adding an "AI feature" to existing products but are directly building AI-native applications or even AI agents that can perform tasks autonomously.

For example, Keystone's "AI engineer" and Solva's "AI claims adjuster" are not just auxiliary tools but "digital employees" that directly replace parts of the workflow. This represents that the moat for AI startups is deepening from "owning data" to "reconstructing processes", and their market substitution potential and commercial value are incomparable.

2: The Rise of Infrastructure and Toolchains

In the AI gold rush, the "water seller" model of providing tools and services to gold diggers remains an investment hotspot and is becoming more vertical and professional. Autumn specializes in the unique "billing" problem in the AI field; Dedalus Labs is committed to lowering the development and deployment threshold for AI agents; Design Arena solves the "quality assessment" problem of AI-generated content.

Investing in such companies is equivalent to investing in the entire AI track. No matter how the upper-layer applications iterate, as long as the AI economy continues to grow, these infrastructure providers will continue to benefit.

3: "Verticalization" and "Necessity" of AI Application Scenarios

AI is accelerating its entry from general chat, image generation, and other fields into more vertical industries with higher commercial value, solving specific and high-cost pain points. Whether it's Solva's entry into the traditional insurance industry, Pingo AI's transformation of the language education industry, or Perseus Defense's targeting of the defense industry, they all have in common: clear business models and extremely strong customer willingness to pay. Investors are increasingly valuing the strong commercial execution ability like Solva, which can generate considerable revenue (ARR) in a short period after launch.

4: The Great Potential of "Non-Mainstream" Tracks and Model Innovations

Although AI is the absolute protagonist, Getasap Asia, one of the highest-valued projects at this YC, is a typical case of "transforming traditional industries with technology". And RealRoots, which has generated amazing revenue, proves that business model innovations that address real human needs (social interaction) can unleash great commercial energy even if the technology is not very "hardcore". This reminds us that investment perspectives need to go beyond simple technology worship and return to the essence of business.

Conclusion

These trends reveal the investment logic behind the public information, but real decision-making requires answering deeper questions:

How deep is Keystone's technological moat?

What is Solva's customer acquisition cost, and can its model be scaled?

How will Autumn's open-source model generate profits?

The answers to these questions cannot be found in any public reports but can only come from in-depth conversations with project founders, early investors, or core technology experts.

When your team is arguing about the technical route, when your investment decision is pending, when your product strategy is in the fog... Remember, the confusion you're facing might be the journey that some experts have already overcome. We at Silicon Rabbit believe that real first-hand experience always comes from those who are driving industry change.

This article is from the WeChat official account “Silicon Rabbit” (ID: gh_1faae33d0655), author: Silicon Rabbit. Republished by 36Kr with authorization.