Y Combinator has opened applications for its fall cohort, and one of its startup themes focuses on cutting government waste.
The theme — “Using LLMs Instead Of Government Consulting” — echoes the mission of the Department of Government Efficiency, a Trump administration initiative to slash spending at the federal level.
While the famed accelerator did not explicitly mention DOGE, YC said on its website that it wants to fund startups that build large language model-powered software to take over the kind of work that consulting giants like Deloitte and Accenture do for the public sector.
“There is political pressure to cut wasteful consulting and spending,” YC’s Gustaf Alstromer wrote. “The US government spends over $100 billion a year on consulting. As you might imagine, this isn’t the most efficient or innovative part of our economy.”
Alstromer said that “LLMs today are so good that they can already do the jobs of many consulting firms,” but startups “can do a lot better” at building software for government use.
He added that YC has already backed companies helping agencies get certified to sell to the government or using AI to ensure the legality of new laws and policies.
Y Combinator did not respond to a request for comment from Business Insider.
The accelerator’s pitch comes as pressure mounts to rein in government spending and fix slow-moving bureaucracies.
Earlier this year, the Trump administration asked the federal government’s highest-paid consulting firms to justify their spending on contracts using language that “a 15-year-old should be able to understand.”
DOGE was, until April, headed by Elon Musk.
That’s just one of six themes outlined in YC’s Fall 2025 “Request for Startups,” a wish list that signals where Silicon Valley could see the next wave of breakout companies
Applications for the accelerator close on Monday evening, Pacific time.
Here’s what else YC is betting on:
AI training for blue-collar workers
YC wants to tackle the talent bottleneck — not in tech, but in the trades.
While the AI race has focused on hiring engineers and researchers, there’s a growing shortage of skilled tradespeople like electricians and welders — the very people needed to build the infrastructure behind AI, including data centers and semiconductor fabs.
“We think you could use AI to create personalized training programs to get people job-ready in months, not years,” wrote YC’s Harj Taggar. “This is where multimodal AI could create opportunities,” he said.
His examples included a voice assistant that could coach someone through tasks in real time or an AR/VR environment that could simulate the work, with vision models giving live feedback like a human trainer would.
Video generation as a core technology
YC said video is evolving into a “new basic building block for software.” It will unlock new kinds of apps, developer platforms, and real-time experiences.
“Video generation models are getting really good,” wrote YC’s David Lieb.
“Soon, you’ll be able to generate near-perfect footage of anything, on the fly, for a marginal cost approaching 0,” he said. “When this happens, a lot of new ideas become possible.”
AI-generated video is “definitely going to change media and entertainment” and reshape how we shop and build games and simulations, he added.
The first 10-person, $100 billion company
The first 10-person, $100 billion company? YC thinks it’s not only possible but imminent.
The accelerator said it’s now possible for small, high-agency teams — or even a solo founder — to build a multibillion-dollar company with just $500,000 in seed funding.
New AI tools will “make it easier for ambitious founders to scale with far fewer people,” YC’s Aaron Epstein wrote.
“With smaller, efficient teams at scale, they won’t get bogged down with the politics, excessive meetings, and lack of focus that grinds huge companies to a halt,” Epstein wrote. “They can just focus on winning with better speed and execution.”
The best startups in the future will “optimize for one metric: revenue per employee.”
YC’s focus comes after at least 12 AI startups crossed the $1 billion valuation threshold with small teams, Business Insider reported in May.
“We’re going to see 10-person companies with billion-dollar valuations pretty soon,” OpenAI CEO Sam Altman said in February 2024. “In my little group chat with my tech CEO friends, there’s this betting pool for the first year there is a one-person billion-dollar company, which would’ve been unimaginable without AI. And now [it] will happen.”
Infrastructure for multi-agent systems
A future where AI agents don’t just work alone — they collaborate.
AI agents are evolving from “single-threaded loops” into “multi-agent systems,” YC’s Pete Koomen wrote. These systems can break down big tasks, run in parallel, and tackle complex jobs faster than any single model could
But they are difficult to build and could introduce new problems that need to be solved at a “higher level of abstraction,” Koomen said.
YC said it’s looking for builders who have “felt this pain in production” and want to create tools to make multi-agent systems easier to build and maintain.
AI-native enterprise software
Lastly, YC said it sees a massive opportunity in reinventing enterprise tools from the ground up, with AI at the core.
Just like how Salesforce and ServiceNow rode the cloud wave 25 years ago, YC said the next generation of tech giants will be built around AI-native software.
“Today’s incumbents will struggle to rebuild their product around this new technology, giving today’s startups the time they need to win,” said YC’s Andrew Miklas.
These startups will have AI “embedded deeply and thoughtfully throughout, and will help employees do their work faster and more accurately,” he said. “Think Cursor for sales, HR, and accounting.”
Read the full article here
