Contract Data in the Age of AI: Why MCP Changes Everything
By Martin Ertl, CEO & Founder of Recital.
Bill Dobie’s insight into maritime data also exposes exactly what’s broken in contract management. In his LinkedIn post about using Claude to analyze shipping data in 30 seconds, he notes: “the need for ‘specialised’ tools becomes unnecessary when source information can be accurately ingested into general models.”
The real value comes from the data.
The Contract Data Problems
Contract data has three persistent problems:
It’s scattered.
Small businesses lose contract docs in email threads or save them inconsistently—if at all—across cloud storage, desktops, and network drives. Larger companies have the same problem, just scaled: data spread across CRMs like Salesforce, legacy CLMs (sometimes multiple), and other systems.
It’s expensive to organize.
Most CLM tools require heavy manual data extraction, custom integrations, and months-long implementations. Six to twelve months. Six to seven figures. ROI? Years away.
It’s locked down.
Many legacy CLMs keep your data behind proprietary APIs and restrictive terms. If you want to use Claude or GPT-4 on your own contracts—or move to a better tool—you can’t. Your data is trapped.
All the while, years of valuable contract history—negotiation emails, redlines, deal evolution—sits untapped. These documents reveal pricing patterns, risk preferences, and negotiation strategies. But they’re buried in email, storage drives, CRMs, ERPs, and outdated CLMs.
This is where the Model Context Protocol (MCP) changes everything.
The MCP Solution
Think of MCP as a universal translator between your contract data and any AI model. Instead of being tied to a vendor’s limited AI, you can connect your entire contract repository—including historical negotiations—directly to Claude, ChatGPT, or whatever works best for your business.
Why be locked into one tool’s limited AI when you can apply any model you want?
As Bill asks: Do you really need a specific vendor’s AI when modern models can understand increasingly complex domains?
Data Flow and Data Portability, Not Data Hoarding
I believe the future isn’t about proprietary contract AI. It’s about letting contract data flow freely—to any model, and into any system that needs it.
That means:
MCP-ready architecture that connects to any model in minutes
Open APIs that let contract data flow to key business apps—ERP, CRM, finance, BI—without ever locking you out of your own data
The future of contract intelligence isn't better CLM platforms—it's liberated contract data working with whatever AI serves your business best.
That said, there's still value in specialized contract intelligence built on top of foundational models—combining the power of clean data with domain expertise. That's the direction contract management needs to go in.
