Are Nscale And Other AI Startups Outpacing Economic Reality?
Two years ago, Nscale didn't exist. Today, it's securing $6.2 billion contracts with Microsoft and collecting $683 million equity checks from...
3 min read
Writing Team
:
Mar 17, 2026 8:00:01 AM
Private equity firms routinely spend between $500,000 and $1 million hiring McKinsey, Bain, or BCG to interview corporate customers and produce the kind of 200-page commercial research reports that underpin billion-dollar acquisition decisions. And if the deal falls through — which deals frequently do — that money is gone. No reimbursement. No credit toward the next deal. Just sunk cost.
That structure creates a perverse incentive: PE firms wait until they're nearly certain about a target before commissioning the research that would tell them whether they should be certain. The diligence that should inform conviction gets delayed until after conviction already exists.
DiligenceSquared, a YC Fall 2025 startup, just raised a $5 million seed round to fix that — and the founding team is credentialed enough to be taken seriously.
Co-founder Frederik Hansen was a principal at Blackstone, where he personally commissioned commercial diligence reports for multiple billion-dollar buyouts. Co-founder Søren Biltoft spent seven years leading exactly these kinds of diligence engagements in BCG's private equity practice. The third co-founder, Harshil Rastogi, is a former Google engineer.
This is not a team that stumbled into a market they don't understand. These are people who have sat on both sides of the table — commissioning and producing the research — and built a product specifically to address the inefficiencies they observed repeatedly in practice. That domain depth is what separates DiligenceSquared from a generic AI research tool with a PE-flavored pitch deck.
The $5 million seed round was led by Damir Becirovic, a former Index Ventures partner, out of his new firm Relentless. Early traction with several of the world's largest PE firms and mid-market funds preceded the raise — the fundraise followed proof of concept, not the other way around.
Instead of deploying expensive management consultants to interview corporate customers, DiligenceSquared uses AI voice agents to conduct those interviews at scale. The approach mirrors what consumer research startups like Listen Labs — which raised $69 million at a $500 million valuation in January — have built for market research. But DiligenceSquared argues the application and output are fundamentally different.
Consumer research synthesizes broad sentiment. Commercial due diligence for a PE acquisition requires C-suite-level interviews, proprietary market data, and outputs that can withstand the scrutiny of an investment committee evaluating a deal worth hundreds of millions of dollars. DiligenceSquared addresses that quality bar by keeping senior human consultants in the loop to verify accuracy and commercial insight in the final output. The AI does the groundwork. Experienced humans validate what it produces.
The result: an analysis that would cost $500,000 to $1 million through a traditional firm was delivered for $50,000. A 90-plus percent cost reduction while maintaining the quality bar that institutional investors actually require.
The more interesting consequence of that cost reduction isn't the savings on any individual project. It's what becomes possible when the price barrier moves.
At $500,000 to $1 million, commercial diligence is a late-stage commitment — something you authorize when conviction is already high, and the deal is likely to close. At $50,000, it becomes an early-stage tool. PE firms can now commission meaningful commercial research before they've committed to a target, using the findings to inform whether conviction is warranted rather than to confirm conviction that already exists. That's the right sequence, and it's only accessible when the price makes early engagement rational.
Hansen puts it plainly: "We are taking these great insights that were previously reserved for the very big decisions, and now we make them more accessible." Accessibility here doesn't mean democratizing research for its own sake. It means correcting a structural flaw in how high-stakes decisions get made.
DiligenceSquared is not operating alone. Competitor Bridgetown Research raised a $19 million in Series A funding co-led by Accel and Lightspeed in February 2026, signaling that institutional investors see this market as real and competitive. The race to own AI-assisted due diligence is underway.
For marketing and growth leaders tracking where AI is creating genuine structural value rather than incremental convenience, this category is instructive. The most durable AI applications aren't the ones that make existing processes marginally faster. They're the ones that make previously inaccessible capabilities accessible — changing who can afford to do something well, and when in a decision process, they can afford to do it.
The management consulting industry has operated on information asymmetry and access barriers for decades. AI voice agents conducting institutional-grade interviews at a fraction of the cost is a direct challenge to that model — not because the technology is flashy, but because the economics are finally right.
For businesses thinking about where AI creates real leverage in their operations, DiligenceSquared offers a useful frame: identify the process where cost or complexity artificially delays a decision that should happen earlier, and ask whether AI can move that threshold. The answer, increasingly, is yes.
If you want to identify where AI can create that kind of structural leverage in your marketing and growth operations, Winsome Marketing's strategists can help you find the right problems before you build toward the wrong solutions.
Two years ago, Nscale didn't exist. Today, it's securing $6.2 billion contracts with Microsoft and collecting $683 million equity checks from...
1 min read
Frank Cilluffo didn't mince words on The POWER Podcast: "If we want to be AI dominant, we can't do that if we're not energy dominant. The two are...
ByteDance released Seed2.0 on February 14th, and the story is becoming uncomfortably repetitive: Chinese AI models matching Western competitors on...