Crunched
Crunched helps advisory teams automate high-stakes financial work inside the tools they already use. Its AI analyst can build models, extract data, check workbooks, generate outputs, and help teams move faster without leaving Excel or PowerPoint.
The product is used across serious advisory and finance workflows, including teams at CBRE, Mile Marker Advisors, Flanders Investment Company, and DW Real Estate, plus a collaboration with S&P Capital IQ. These are firms working on transaction advisory, commercial real estate, private equity, and investment analysis, where speed matters but the numbers still have to hold up.
As Crunched expanded across customers, one request kept coming back: can the AI analyst work with our internal systems?
That meant SharePoint, Google Drive, data rooms, financial datasets, ERP systems, company registries, and private customer tools. Each one came with its own API, OAuth flow, scopes, permissions, and security review.
Crunched uses Metorial so those customer systems can become governed MCP tools without turning every connector into a custom infrastructure project.
The Challenge
Every customer has a different stack
Crunched’s customers are not asking for a generic AI assistant. They want an AI analyst that can work with the documents, models, research sources, and systems behind their actual advisory work.
That creates a practical problem. The more customers Crunched serves, the more systems the product needs to reach.
Building every connector internally would mean owning the same infrastructure over and over again: OAuth, scopes, refresh tokens, tenant rules, tool definitions, MCP servers, and permission boundaries.
The team needed a way to support customer-specific systems without becoming a connector company.
The Solution
Metorial handles the integration layer
Metorial turns approved customer systems into tools Crunched’s AI analyst can use. It handles the authentication, permission scoping, and MCP surface underneath.
That lets Crunched keep its product centered on what customers buy it for: financial analysis, Excel-native workflows, PowerPoint outputs, review, and verification.
The split is simple: Crunched builds the AI analyst. Metorial connects it to the customer’s tools.
Focus Area
Document systems
SharePoint, Google Drive, and data rooms hold much of the source material behind advisory workflows. Metorial gives Crunched a way to expose those systems with access tied to the user.
Financial and market data
Customers often need financial datasets, company registries, or industry-specific sources connected to the workflow. Metorial reduces the custom work needed to support each source.
Private customer tools
Some systems are specific to one customer. Metorial gives Crunched a repeatable path for exposing them without creating a one-off architecture every time.
Permissions and governance
Customers need to know what the AI analyst can access. Metorial handles OAuth, scopes, delegated permissions, and per-user access so connectors can be rolled out with control.
The Result
More customer workflows, less connector work
With Metorial, Crunched can respond to customer connector requests faster and with less infrastructure burden.
Customers get an AI analyst that works closer to their real environment: documents, data sources, internal systems, and approved tools. Crunched keeps its engineering team focused on the workflows that make the product valuable.
For advisory work, that matters. The quality of the output depends on the quality of the input. Metorial helps Crunched bring the right data into the workflow without rebuilding the access layer for every customer.
Broader Impact
The AI analyst needs access to the work
The next step for AI in advisory is not a better chat window. It is giving AI systems safe access to the tools and data where the work already lives.
Crunched brings the analyst experience. Metorial brings the connector layer. Together, they let customers use AI on real financial workflows without turning every new system into a custom infrastructure project.