Crunched transforms Excel modeling for investors and advisors with Claude

Crunched is an AI Excel analyst built for Excel power users, serving management consulting, investment banking, private equity, and investment firms. 

With Claude, Crunched:

• Delivers up to 50% time savings across Excel modeling tasks for investor and advisor workflows, by leveraging Sonnet and Opus 4.5
• Error-checks models of any size, catching linking mistakes, logical flaws, and finance-specific methodology errors
• Enables autonomous model building and template linking that delivers up to 10x speed improvements on complex workflows
• Integrates data from PDFs, images, web sources, third party financial data providers, and customer-specific databases

The problem

Consulting and finance professionals working on high-stakes deals face a persistent challenge: Excel modeling is manual, error-prone, and slow. Private equity associates spend days building financial models from investment memorandums under tight deadlines. Investment banking analysts waste iteration cycles when VPs catch linking errors and logical flaws in first-pass reviews. Consultants building market sizing models or business cases need granular, verifiable assumptions but lack time to gather and structure data from multiple sources.

Philip Borge and Michael Sakowski built Crunched after spending 10,000+ hours in Excel at McKinsey's private equity practice experiencing these exact problems. For the top 1% of Excel users worldwide, models are either the main deliverable or the backbone of every presentation. These professionals needed speed and accuracy to complete more deals, dive deeper into analysis, and deliver outputs their clients could trust.

"Excel modeling is too slow and too error-prone," said Borge, CEO and co-founder. "We were solving our own problem, and we knew millions of other finance and consulting professionals faced the same challenges."

Claude enables high-speed, accurate modeling for power users

Crunched needed an AI foundation that could handle complex Excel models—workbooks with 50+ tabs and thousands of formulas—while maintaining accuracy on finance-specific calculations like DCF and LBO modeling. Before Sonnet 4.5, they built a sophisticated context searching and management system, their core differentiator, that gathers precisely the right information for each task, enabling accurate operations on workbooks of any size.

Sonnet 4.5 dramatically extended what that system could do. The model's parallel tool calling lets Crunched gather context from multiple workbook locations simultaneously, while its extended context window allows the system to process more information at once, reducing the risk of missing critical dependencies. Low latency enables the real-time iteration speed that power users demand. "Sonnet 4.5 is the best booster of our context searching system—it increased its range by a lot," said Sakowski.

Crunched builds financial and Excel-specific knowledge on top of Claude's foundation, augmenting the model with deep domain expertise in Excel modeling best practices, finance methodologies, and consulting frameworks. They evaluated multiple LLMs before choosing Claude for its combination of speed, accuracy on quantitative tasks, and reliable performance on complex, multi-step financial modeling workflows.

Crunched brings AI-powered Excel analysis to finance workflows

Backed by Y Combinator and First Round Capital, Crunched took a different approach to AI implementation. Rather than building a generic tool, the company focused on specific pain points for management consulting, investment banking, private equity, and real estate firms. The team works on-site with customers, tailoring Crunched's capabilities to industry-specific modeling workflows, company-specific template tasks and formatting, and regional data source integrations.

"Driving Excel productivity is not only about building models from scratch" Sakowski explained. "Professionals aren't asking AI to create another DCF template—they already have those. They need help solving firm-specific pain points one at a time to drive productivity."

The forward-deployed model means Crunched solves concrete problems. For understanding and error-checking models, Crunched analyzes models of any size, identifying everything from linking mistakes to logical flaws and methodology errors specific to finance. Customers use the error-check as a review layer between analysts and VPs, where Crunched handles the first-pass model review. This saves iteration cycles, reduces embarrassing moments for analysts, and gives VPs higher confidence in the work they receive.

For financial model building and editing, Crunched fills out templates, links into input sheets, in addition to building out valuation models, business cases and analyses from scratch. When a private equity associate receives a counterpart's model and needs to link a "management scenario" into their own LBO model, Crunched handles it. When a consultant needs a fully-linked market model sizing the total addressable market for online beauty products in Florida, Crunched builds it with granular assumptions, verifiable rationales, and clickable links.

"Anthropic allowed us to be more specific in the workflows we tailor to, delivering higher time savings on tasks like building business case projections across all three financial statements for private companies based on Crunched's own research and uploaded investment presentations," Sakowski said.

For data gathering, Crunched pulls information from PDFs, images, web sources, and customer-specific external and internal databases. A private equity associate can upload an information memorandum and build an LBO with granular drivers and assumptions tailored to the specific case, arriving well-prepared for a first meeting. A consultant tasked with reviewing a business model can have Crunched scan all assumptions and assess whether they're optimistic, realistic, or pessimistic, complete with rationales and verifiable links to sources.

The outcome

Customers report over 50% time savings on Excel modeling tasks with Claude Sonnet 4.5, with some workflows showing dramatic improvements: financial spreads that took 8+ hours now take 1 hour, and company write-ups dropped from 8 hours to 20 minutes. Crunched implemented the Claude integration in less than a day.

For an investment banking team, first-pass model reviews that typically required VP oversight now happen in minutes, reducing iteration cycles and accelerating deal timelines. A management consulting firm uses Crunched to build fully-linked market sizing models in hours instead of days, arriving at client meetings with granular, verifiable assumptions that strengthen recommendations.

For one real estate investor, the impact was immediate and measurable. A deal-specific task combining automated model building and web-based company research dropped from 10 hours to 1 hour—a 10x speed improvement.

Transactions happen under tight deadlines with finite resources. Senior dealmakers constantly identify analyses that would be valuable but remain too expensive given time and capacity constraints—the transaction size often doesn't justify allocating more associates to a deal.

With Claude Opus 4.5, Crunched enables this economically useful work that was previously cost-prohibitive. The pattern extends across project and deal types. Private equity teams bring full analytical rigor to early-stage deals that previously only received deep analysis at later stages. Investment bankers run additional scenarios that strengthen client recommendations. The ability to economically pursue inherently valuable analyses earlier and more comprehensively resonates with senior dealmakers across finance and consulting.

“We're fundamentally changing how power users work in Excel,"  Borge said, "Automating all  grunt work, one pain point at a time, tailoring Crunched to each firms' specific needs."