BridgeSuite™ · Verbatim™ · Powered by LitAgility™
Verbatim returns the actual language from your source documents – cited, linked, and immediately verifiable. No paraphrasing. Every citation links to the exact passage in your source document, so you can verify it in seconds.
the problem
Patent litigation requires exact language from exact sources. Whether you’re mapping claim elements to prior art, building an infringement analysis, or preparing deposition testimony for cross, a paraphrase isn’t the same as the actual text – and a citation you can’t verify quickly is a liability.
Most AI tools return fluent summaries. They compress, infer, and occasionally fabricate. As serial tasks accumulate, the same three failure modes compound — Input Overflow, Attention Dilution, and Output Truncation. Verbatim’s managed RAG memory holds each in check across the full workflow.
Verbatim was built for exactly this problem.
“Click any cited passage. Verbatim opens the source PDF, navigates to the page, and puts a red box around the exact text. Verification takes seconds.”
how it works
Verbatim is part of BridgeSuite, LitAgility's document-intelligence suite for patent litigation.
A typical patent litigation session moves between all three – sometimes in the same conversation.
MODE 01
CITATION-PRECISE
Returns exact quoted language from your uploaded documents – no paraphrasing, no inference. Every result is cited by document, page, and location, with a live link for instant verification.
MODE 02
Broad document interrogation for orientation and issue identification. Useful for quickly understanding a large production set or identifying relevant passages before narrowing to verbatim extraction.
MODE 03
Optional web research and reasoning, kept separate from the document-anchored modes so you control when anything outside your case materials is used. Useful for legal research, drafting, and case analysis – integrated in the same workflow.
Verbatim is model agnostic. A single license gives you access to the latest frontier models from Gemini, Claude and ChatGPT — switchable per task. Use the best model for technical claim analysis, a different one for long-document synthesis, another for drafting. All traffic runs through one Azure-hosted matter workspace with a single data agreement and one audit trail. When a better model is released, you can use it immediately – no new procurement, no security reviews, no tool change.
Instant Verification
When Verbatim returns a result, you don’t have to take its word for it. Every cited passage is clickable.
Click any cited passage
in the Verbatim Chat output
Source PDF opens automatically
no manual file hunting
Navigates to the exact page
and highlights the passage in a red box
Deposition of R. Chen — Page 94
Q. And you received that document before the filing date of the patent application?
A. “Yes. We had that in our possession by early March. We circulated it to the product team within the week.”Q. Do you recall who specifically received that circulation?
A. I’d have to check the email records to give you an exact list …
ILLUSTRATIVE EXAMPLE – NOT CASE-SPECIFIC
where it gets used
Not a general tool adapted for IP – a platform designed for this workflow.
Return every passage in the specification and prosecution history bearing on a disputed claim term – verbatim, cited by column and line.
Query accused product documentation against each claim element. Get the relevant passages, not a summary of them.
Load references, patents, and standards. Ask what each one discloses against specific elements. Element-by-element mapping without manual document review.
Find all testimony on a specific issue, or identify contradictions between a witness’s statements. The result is a set of cited verbatim excerpts ready for use.
Query license agreements and financial documents against specific damages theories. Get the relevant terms and provisions verbatim.
Interrogate key exhibits against specific factual claims. No manual searching, no risk of paraphrase in what gets cited.
Evaluate written description support by querying the patent and application directly. Every relevant passage, sourced and cited.
Apply the same query to every document in a set. Verbatim returns per-document results with citation integrity throughout.
Verbatim reviews uploaded source documents to understand the disclosed technology and provides context-aware insights using domain-specific expertise for your attorney/expert to review and finalize. For inventions spanning multiple technical disciplines, the review draws on expertise across relevant subject areas to surface accurate and relevant results.
Analytical consistency
Most AI tools work from a fixed context window. As document volume grows, general AI tools lose fidelity — “token dilution,” which shows up as Input Overflow, Attention Dilution, and Output Truncation. Verbatim managed RAG memory holds each in check across the full workflow.
Serial claim charts
Consistent interpretation of claim terms maintained across large patent families.
multi-reference obviousness
Chain prior art under KSR without confabulating which technical features appear in which reference.
multi-IPR proceedings
Consistent argument framing across petitions covering dozens of patents with overlapping prior art.
prosecution history review
Accurate tracking of claim amendment positions and applicant arguments across lengthy file wrappers.
What’s Different
The difference isn’t just the model. It’s document processing, domain-tuned assistants, and citation integrity working together.
Context window degrades across extended workflows. Token dilution compounds with document volume – models confabulate or conflate content, and users must manually re-establish context for every new step.
Managed RAG memory maintains analytical fidelity across serial tasks. No manual context re-establishment.
The AI returns a fluent paraphrase of what the specification says about claim 3.
You get the exact language from the specification – quoted, cited by column and line, and linked to the source.
You spend 10 minutes locating a passage in the original document to confirm what the AI told you is accurate.
You click the citation. The source PDF opens at the correct page with the passage highlighted. Verification takes seconds.
Reviewing 40 deposition transcripts for a specific theme takes days of paralegal time.
Batch Prompt applies your query to all 40 transcripts sequentially, returning cited per-document results in minutes.
You’re locked to a single AI model regardless of the task – technical claim analysis, long-document synthesis, and drafting all run through the same engine.
One license gives you the latest frontier models from OpenAI, Anthropic, and Google — switchable per task. One Azure workspace. One data agreement. One audit trail. Use the best model for each job.
You research and define the skilled artisan/ POSITA standard manually, then work to apply it consistently across claim construction, obviousness analysis, and expert declarations – a common source of drift.
Verbatim automatically determines the applicable skilled artisan standard from your uploaded patent and prosecution materials – no prompting, no manual research – and applies it consistently to every task.
Security & Data Isolation
Built and operated by LitAgility. Verbatim runs on LitAgility's secure BridgeSuite infrastructure - one dedicated, Azure-hosted workspace per matter, encrypted at rest and in transit, with role-based access controls. Your documents are never used to train any model.
Every matter gets its own dedicated, isolated environment. Documents you upload for one case are never accessible to another matter, another firm, or anyone else. BridgeSuite runs on Microsoft Azure with enterprise-grade controls throughout.
Dedicated workspace per matter
Each case or project operates in its own isolated environment. Documents, queries, and results are scoped to that workspace and never commingled with other matters.
No training on your data
Your documents are never used to train AI models — ours or our providers’. Uploads are used only to answer your queries within your matter workspace.
Hosted on Microsoft Azure
BridgeSuite is built on Azure’s enterprise cloud infrastructure – the same platform used by the world’s largest law firms and financial institutions for sensitive workloads.
Encryption in transit and at rest
All data is encrypted end-to-end. Documents you upload are protected in storage and in transit, with no exposure to shared infrastructure.
Powered by Microsoft Azure
Role-based access controls let you define exactly who on the matter sees what, with a full audit trail of activity.
Simple, Predictable Pricing
No annual subscription
Pay per case/matter.
See It In Action
A 90-second walkthrough of a 39-document invalidity analysis - from initial document orientation to verbatim citation extraction.
— Partner, IP Litigation, AM100 law firm
— Deputy General Counsel, multinational
© 2026 LitAgility, LLC. All rights reserved.
BridgeSuite™ and Verbatim™ are trademarks of LitAgility, LLC.