Overview
Custom Terms can be extracted using Pramata's AI Clause Analysis capabilities. AI Clause Analysis is a powerful Pramata feature that enables you to identify and extract specific clauses and terms from your contracts at scale. Whether you need to understand payment obligations, termination rights, liability caps, or any other contractual concept across your portfolio, Clause Analysis delivers structured, searchable intelligence from your documents.
You may also hear this feature referred to as custom terms or custom term extensions. These names all refer to the same capability: extending your contract data beyond the standard terms included with File Cleanse and Document Organization by defining the specific clauses and terms you want the AI to find and extract.
How AI Clause Analysis Works
Clause Analysis uses a Contract AI Model — a structured taxonomy that tells the AI exactly what to look for in your documents. The model is organized into three hierarchical levels:
| Level | What It Is | Why It Matters |
| Clause | A high-level container grouping related contractual concepts (e.g., “Payment,” “Termination,” “Confidentiality”). Clauses organize your model logically. | Provides the organizational structure for your model. You can have as many Clauses as needed. |
| Term | A specific concept or obligation within a Clause (e.g., “Payment Term,” “Payment Method”). The AI locates the corresponding language in your contract and saves it as a reference point for further analysis. | The core unit of analysis. The number of Terms in your model determines the depth and breadth of extraction. |
| Data Element | The extracted, structured details from a Term (e.g., “Number of Days,” “Currency”). Data Elements capture the specific values you need for reporting and analysis. | Delivers the granular, structured data points that power your reporting and decision-making. |
Here’s a concrete example of how these levels work together:
| Clause | Term | Data Elements | Extracted Data Element Values |
| Payment | Payment Term | Number of Days, Payment Trigger | Number of Days: 30 Payment Trigger: Invoice Date |
| Payment | Payment Method | Currency | Currency: USD |
| Termination | Termination for Convenience | Notice Period, Effective Date | Notice Period: 90 days Effective Date: Written notice |
In this example, the Payment Clause contains two Terms (Payment Term and Payment Method), each with their own Data Elements. The AI finds the relevant language in your contracts, extracts the structured data, and makes it searchable across your entire portfolio.
Clause Analysis is powered by Contract AI Credits (CACs) — Pramata’s consumption-based currency that covers platform features including Clause Analysis. The number of Terms in your Contract AI Model is the primary driver of how credits are consumed. Your Pramata Account Manager can provide full details on how CACs apply to your specific agreement and help you plan effectively.
Get Started Quickly with the Pramata Starter Model
You don’t need to build a Contract AI Model from scratch. Pramata provides a Starter Model with approximately 50 pre-defined Terms covering the most commonly analyzed contract Clauses and Terms. This out-of-the-box model is ready to use immediately and gives you a broad foundation of contract intelligence from day one.
What’s in the Starter Model?
The Starter Model includes Clauses and Terms such as Payment Terms, Termination Rights, Auto-Renewal, Liability Limitations, Indemnification, Confidentiality, Governing Law, Assignment, Insurance Requirements, and more. Contact your Pramata Account Manager for the full list of included Clauses and Terms.
You have three ways to leverage the Starter Model:
- Use it as-is. Run the full Starter Model against your contracts to extract all ~50 Terms. This is the fastest way to get comprehensive contract intelligence.
- Modify it. Start with the Starter Model, then add, remove, or redefine Clauses, Terms, or Data Elements to match your organization’s specific needs. This is the most popular approach — it gives you a solid foundation while allowing full customization.
- Build a fully custom model. If your contracts have unique Clauses and Terms not covered by the Starter Model, you can define entirely custom Clauses, Terms, and Data Elements from scratch.
What You Can Do with Clause Analysis
Once your Contract AI Model is configured, the possibilities are broad. Here are some of the most common use cases our customers pursue:
- Risk and compliance reviews. Quickly identify contracts with unfavorable liability caps, missing indemnification clauses, or non-standard governing law across thousands of agreements.
- Renewal preparation. Extract auto-renewal terms, notice periods, and pricing escalation clauses so your team has a clear picture before every renewal conversation.
- M&A due diligence. Analyze an entire portfolio of acquired contracts to surface change-of-control clauses, assignment restrictions, and termination triggers.
- Standardization initiatives. Compare actual contract language against your organization’s standard playbook to identify deviations and prioritize remediation.
- Financial and operational reporting. Extract payment terms, pricing structures, and service levels to feed downstream reporting and analytics.
Tailor Your Analysis to Different Contract Types
One of the most powerful aspects of Clause Analysis is that you don’t have to run the same model against every document. You can apply different Contract AI Models — or different subsets of Terms — to different types of contracts. This ensures each document gets the right depth of analysis for its purpose.
For example:
- MSAs and Amendments: Run the full Starter Model to extract comprehensive commercial and legal Terms — including termination rights, liability caps, indemnification, and governing law. Amendments can be analyzed with the same model to capture how key Terms have changed over time.
- Order Forms: Use a targeted subset of Terms focused on pricing, product entitlements, and renewal dates.
- Buy-Side vs. Sell-Side Agreements: Different sides of the table often call for different analysis. For buy-side contracts, you might focus on Terms like payment obligations, service levels, and warranty coverage. For sell-side agreements, the priority may shift to Terms like revenue recognition triggers, limitation of liability, and intellectual property ownership. You can build separate Contract AI Models — or tailor subsets of the Starter Model — to match the Terms that matter most for each.
This flexible approach means you can focus your analysis where it delivers the most value.
Best Practices for Getting the Most from Clause Analysis
- Start with the Starter Model. The pre-built ~50 Term model covers the most common Clauses and Terms and gives you immediate value. Customize from there based on what your organization needs.
- Test on a sample first. Before running your model across your entire portfolio, test it on a small sample of contracts to verify the results are accurate and the model captures what you expect.
- Right-size your model. Focus on the Terms that deliver real business value. You can always add more Terms later and reprocess specific document subsets as your needs evolve.
- Follow Pramata’s design guidelines. Well-designed models produce better extraction accuracy. Your Pramata team can help you optimize your model design for the best results.
- Segment your portfolio. Not every document needs the same depth of analysis. Group your contracts by type and apply the appropriate model to each group for the most effective results.
Frequently Asked Questions
Q: Can I use Clause Analysis on any type of contract?
A: Yes. Clause Analysis is contract-type agnostic. You can apply it to procurement contracts, sales agreements, NDAs, leases, or any other contract type in your portfolio.
Q: Do I need to build a model from scratch?
A: No. The Pramata Starter Model gives you approximately 50 ready-to-use Terms covering the most common Clauses. You can use it as-is, customize it, or build your own from scratch — whichever approach fits your needs.
Q: What happens if I want to change my model after I’ve already processed documents?
A: You can update your Contract AI Model at any time. If you add new Terms or modify existing ones, you can reprocess specific document subsets with the updated model to capture the new information.
Q: How long does Clause Analysis take?
A: For batch runs of more than 1,000 documents, results are typically available within 2 to 24 hours. For larger batches exceeding 15,000 documents, processing may take longer than 24 hours.
Q: How does Clause Analysis affect my credits or billing?
A: Credit consumption for Clause Analysis depends on your specific agreement with Pramata. For details on how Clause Analysis applies to your account, please reach out to your Pramata Account Manager — they can walk you through the options and help you plan.
Ready to Get Started?
AI Clause Analysis can transform how your organization understands and manages its contract portfolio. Whether you’re looking to surface risk, streamline renewals, or build a searchable knowledge base of contract obligations, the Starter Model gives you a fast, proven starting point.
Reach out to your Pramata Account Manager to discuss how Clause Analysis can work for your organization, explore the Starter Model, and plan your next steps. You can also contact us anytime at support@pramata.com.