An eight-hour financial model is a billable day. Crunched, a New York-based startup founded last year, sells the promise of turning that day into an hour. It’s a bet that the world’s most entrenched financial analysis tool, Microsoft Excel, is not the problem. The problem is the human hours spent populating, debugging, and iterating on the spreadsheets inside it.
Led by former McKinsey Associate Partner Philip Golub and co-founder Matthew Reims, the company has raised $6 million in seed funding to build what it calls an "Excel-native AI analyst." The tool runs as a cloud-based add-in, promising to check workbooks for errors, pull data from web searches and PDFs like annual reports, and build analyses from scratch. The target buyer is the ambitious advisor in consulting, investment banking, or private equity, the kind of professional for whom Excel is a second language and time is a direct revenue metric.
First Round Capital led the round, with participation from Y Combinator, 20VC, Founder Factor, and Multimodal Ventures [Crunched company blog, 2024]. The backing from First Round, a firm with a deep bench of enterprise software investments, signals a belief that the wedge into Excel is a defensible one.
The Wedge Is the Workspace
Crunched’s core proposition is not a new application. It is an agent that lives inside an existing one. The product integrates directly into Microsoft Excel, positioning itself as a co-pilot that understands financial modeling logic and can operate within a user’s existing workbook [Y Combinator, 2024]. This is a deliberate choice against the grain of many AI productivity tools, which ask users to leave their primary work surface for a new web dashboard or chat interface.
For the target user, the friction of context-switching is a real cost. "The first AI Excel agent built by and for Excel power users," the company claims, is designed to let teams "review and iterate as usual" without exporting or reformatting data [Y Combinator (LinkedIn post), Oct 2024]. The workflow promises are specific: compressing one-hour modeling tasks into 5-10 minute automated sequences, and turning company write-ups that took eight hours into 20-minute exercises.
The Founder's Edge
The company’s trajectory is tightly linked to its CEO’s background. Philip Golub spent years as an Associate Partner at McKinsey & Company, focusing on financial institutions and private equity, before earning an MBA from Stanford Graduate School of Business [LinkedIn, retrieved 2024]. That experience maps directly onto Crunched’s ideal customer profile. Co-founder and COO Matthew Reims brings operational experience from a prior role at Osprey Accounting [ZoomInfo, retrieved 2026].
This isn’t a team building a general-purpose AI tool and looking for a market. It’s a team that emerged from the market they’re now selling to. The product roadmap,error-checking, data extraction, bespoke analysis,reads like a list of pain points from a consultant’s late-night spreadsheet review. Golub’s network within elite advisory firms provides a natural early-adopter channel and a source of product feedback that is difficult for outsiders to replicate.
Where the Model Could Stumble
Embedding inside Excel is a clever distribution strategy, but it also creates dependencies. The tool’s utility is contingent on the stability and API accessibility of Microsoft’s ecosystem. While Excel is a monolith unlikely to be displaced, changes to its architecture or licensing could introduce complexity. Furthermore, the "AI analyst" label invites high expectations for reasoning and accuracy in complex financial scenarios. A single material error in a sensitive model could damage trust more profoundly than a mistake in a less critical document.
The competitive landscape, while not crowded with direct clones, is adjacent. Large consulting firms and banks have internal automation teams. Microsoft itself is aggressively integrating Copilot across its Office suite. Crunched’s answer is specificity: a tool built by finance professionals for finance workflows, claiming a depth of domain understanding that a generalist assistant cannot match. Early traction signals, including user-reported time savings of over 50% on modeling tasks, suggest the product is finding its fit.
The company’s current investor lineup and reported customer enthusiasm point to a focused execution path. The key milestones to watch will be less about flashy new features and more about commercial proof:
- Enterprise penetration. Moving from individual power users to firm-wide deployments at named consulting or banking clients.
- Renewal motion. Demonstrating that the time savings translate into sticky, high-value subscriptions, not just one-off project use.
- Expansion beyond modeling. The product already mentions AI-powered presentation review [review.usecrunched.com, retrieved 2024]; successfully layering adjacent tasks would increase its share of wallet.
The Funding and the Road Ahead
The $6 million seed round provides runway to refine the product and scale initial sales efforts. The involvement of Y Combinator (F25 batch) and a syndicate that includes 20VC and Founder Factor suggests confidence in both the team and the market timing. There is a palpable fatigue with AI tools that create more work; a tool that reduces work inside an existing, beloved (or at least tolerated) environment has a clear narrative.
For Crunched, the next twelve months are about moving from a promising wedge to a measurable beachhead. The goal is not to replace the financial analyst, but to make them exponentially faster. If the company can prove its tool scales across an organization, providing use from intern to director as it claims [Microsoft Marketplace, 2026], the total addressable market stretches to every professional services firm that runs on spreadsheets.
The seed check from First Round Capital values that ambition. The question for Golub and Reims is whether they can turn the consultant’s eight-hour grind into a one-hour task often enough to build a category-defining company around it.
Sources
- [Y Combinator, 2024] Crunched: Excel AI analyst for Power Users | https://www.ycombinator.com/companies/crunched
- [Y Combinator (LinkedIn post), Oct 2024] Crunched is the first AI Excel agent built by and for Excel power users | https://www.linkedin.com/posts/y-combinator_crunchedis-the-first-ai-excel-agent-built-activity-7386053314884534272-wOMh
- [Crunched company blog, 2024] Crunched raises $6M to scale the AI analyst Excel power users deserve | https://www.usecrunched.com/blog/seed-round
- [LinkedIn, retrieved 2024] Philip Golub Profile | https://www.linkedin.com/in/philip-golub/
- [ZoomInfo, retrieved 2026] Matthew Reims Profile | https://www.zoominfo.com/p/Matthew-Reims/-1786731186
- [Microsoft Marketplace, 2026] Crunched | https://marketplace.microsoft.com/en-us/product/saas/wa200009174?tab=overview
- Crunched transforms Excel modeling for investors and advisors | https://claude.com/customers/crunched
- [review.usecrunched.com, retrieved 2024] Crunched Product Page | https://review.usecrunched.com