Y Combinator hosted its Summer 2025 Demo Day last week, showcasing the latest batch of over 160 startups.<\/p>\n
As with recent batches, the majority of startups presented AI-centric solutions. However, a clear evolution was evident. Instead of\u00a0 \u201cAI-powered\u201d products, many companies are now building AI agents or the infrastructure and tools needed to develop them. For instance, this batch had a flurry of voice AI solutions and new businesses focused on helping others monetize the \u201cAI economy\u201d with ads and marketing tools.<\/p>\n
We spoke with a handful of YC-focused\u00a0 investors to learn which startups they found most interesting and which generated the highest investment demand.<\/p>\n
Below are the most frequently mentioned ones:<\/p>\n
What it does: Stripe for AI startups<\/p>\n
Why it\u2019s a fave: Many AI startups use complex pricing models that often blend a flat subscription fee per seat with usage-based charges, credits, and various add-on costs. Managing complex AI pricing on Stripe is a time-consuming, manual process. That\u2019s why Autumn developed an open-source infrastructure that simplifies Stripe integration for AI startups. The company says its technology is already used by hundreds of AI apps and 40 YC startups. Given Stripe\u2019s dominance in payments and the explosive growth of the AI market, could a specialized billing solution for AI be the next major fintech success story?<\/p>\n
\u00a0<\/p>\n
Techcrunch event<\/p>\n
\n\t\t\t\t\t\t\t\t\tSan Francisco
\n\t\t\t\t\t\t\t\t\t\t\t\t\t|
\n\t\t\t\t\t\t\t\t\t\t\t\t\tOctober 27-29, 2025\n\t\t\t\t\t\t\t<\/p>\n
What it does: Builds Vercel for AI agents<\/p>\n
Why it\u2019s a fave: Just as Vercel helps developers deploy and host startups, Dedalus Labs claims its platform automates the infrastructure for building AI agents, cutting hours of coding down to a few clicks. The company handles complex tasks like autoscaling and load balancing, which it says makes agent deployment fast and simple.<\/p>\n
What it does: crowdsource rankings of vibe coded designs<\/p>\n
Why it\u2019s a fave: The ability of AI to rapidly generate a huge number of designs creates a new problem: figuring out which ones are actually good. Design Arena solves this by crowdsourcing rankings of AI-generated visuals, creating a feedback loop that forces AI models to improve. Large AI labs see value in training their models to generate better designs, as some of them are already Design Arena\u2019s customers.<\/p>\n
What it does: Tech-enabled distributor for retailers in Southeast Asia<\/p>\n
Why it\u2019s a fave: Getasap Asia was founded by Raghav Arora three years ago when he was just 14 years old. Since then, the startup that uses tech to deliver supplies to corner stores, restaurants and large supermarkets in Southeast Asia in under eight hours, has earned millions in revenue. Getasap Asia closed a round from General Catalyst, according to its website<\/a>, and we are hearing that the startup\u2019s valuation was among the highest in the whole batch.<\/p>\n What it does: AI engineer that fixes bugs in production<\/p>\n Why it\u2019s a fave: Founded by a 20-year-old Pablo Hansen who last year earned a master\u2019s degree in AI, Keystone is on a mission to reduce software breaks. The company\u2019s AI finds and fixes bugs for clients like Lovable and has already turned down a seven-figure acquisition offer, Hansen said.\u00a0<\/p>\n What it does: AI matchmaker for female friends<\/p>\n Why it\u2019s a fave: While there isn\u2019t a shortage of dating apps, RealRoots is tackling a different kind of loneliness. The company\u2019s AI matchmaker, Lisa, interviews women and then organizes social experiences to connect them with compatible friends. While the AI part might be performative \u2013\u00a0 conversations with Lisa probably wouldn\u2019t give RealRoots more insights about participants than written answers would \u2013 RealRoots may be on to something. Last month alone, the company generated $782,000 from 9,000 paying clients, its founders said.<\/p>\n What it does: Automates insurance claims with AI<\/p>\n Why it\u2019s a fave: Solva\u2019s AI automates the most routine tasks for insurance adjusters, from filling out complex claims to preventing improper payouts. Just ten weeks after launching, Solva has already amassed $245,000 in annual recurring revenue (ARR), a figure that has investors excited.<\/p>\n What it does: counter-drone mini-missiles<\/p>\n