Article

2026-08-04

Drafting discovery and motions from the full matter file

Drafting discovery and motions from the full matter file

Turbo Law connects drafting to the full litigation matter: the evolving Fact Graph, firm prior work product, and jurisdiction rules. Motions and discovery begin with what the record actually says, keep prior answers and admissions in view, and leave factual propositions open to staff verification before a partner decides what gets filed or served.

Turbo Law connects drafting to the full litigation matter: the evolving Fact Graph, firm prior work product, and jurisdiction rules. Motions and discovery begin with what the record actually says, keep prior answers and admissions in view, and leave factual propositions open to staff verification before a partner decides what gets filed or served.

Turbo Law connects drafting to the full litigation matter: the evolving Fact Graph, firm prior work product, and jurisdiction rules. Motions and discovery begin with what the record actually says, keep prior answers and admissions in view, and leave factual propositions open to staff verification before a partner decides what gets filed or served.

An associate preparing a summary judgment motion had a strong draft by Friday afternoon. The argument was organized. The citations looked complete. The factual section followed the firm's preferred style.

Two days later, a privilege review uncovered a problem. The draft relied on a notice date that appeared in an early case summary, but the client's March interrogatory answer gave a different date, and a produced custodian email created another point of tension. The paragraph read well. The problem was that the drafting process never treated the matter file as the source of truth.

Turbo Law approaches litigation drafting from the opposite direction. We connect drafting work to the underlying matter record, so motions and discovery responses begin with the facts, documents, prior work product, and rules that shape the case.

The quality of a litigation draft depends on what the drafting system actually knows about the case. Turbo Law starts with the matter itself, not a blank prompt or a generic template. We connect the evolving case record through an intelligent Fact Graph, then use that matter intelligence alongside firm prior work product and jurisdiction rules to draft motions, discovery responses, and other litigation work product.

That distinction matters. A polished paragraph is not useful if it misstates the record, conflicts with an admission made three months earlier, or ignores the way the firm actually argues the issue.

Drafting is downstream of understanding the case

Litigation drafting is rarely a standalone writing exercise. A motion for summary judgment depends on the evidence supporting each material fact. A motion in limine depends on what the opposing party intends to introduce and why it matters. A motion to compel depends on the requests, responses, objections, correspondence, governing rules, and history of the dispute. Discovery responses can create admissions that follow the client through depositions, dispositive motions, mediation, and trial.

The draft therefore needs access to more than documents. It needs an understanding of the matter. That is the role of our Fact Graph.

The Fact Graph goes beyond entity extraction. It is an intelligent representation of the case that connects the facts, events, people, issues, pleadings, discovery, case law, and case-specific materials that make the litigation understandable. It is also dynamic. The shape of the graph changes by practice area because the facts that matter in a medical malpractice case differ from those that matter in construction, employment, product liability, or a business dispute. It changes again for each individual matter. And it continues changing as complaints are amended, discovery is exchanged, records arrive, depositions are taken, experts enter the case, and new evidence changes the significance of what came before.

That evolving intelligence layer feeds drafting. The motion is not being written next to the case file. It is being written from a connected understanding of the case file.

Drafting is downstream of understanding the case

Litigation drafting is rarely a standalone writing exercise. A motion for summary judgment depends on the evidence supporting each material fact. A motion in limine depends on what the opposing party intends to introduce and why it matters. A motion to compel depends on the requests, responses, objections, correspondence, governing rules, and history of the dispute. Discovery responses can create admissions that follow the client through depositions, dispositive motions, mediation, and trial.

The draft therefore needs access to more than documents. It needs an understanding of the matter. That is the role of our Fact Graph.

The Fact Graph goes beyond entity extraction. It is an intelligent representation of the case that connects the facts, events, people, issues, pleadings, discovery, case law, and case-specific materials that make the litigation understandable. It is also dynamic. The shape of the graph changes by practice area because the facts that matter in a medical malpractice case differ from those that matter in construction, employment, product liability, or a business dispute. It changes again for each individual matter. And it continues changing as complaints are amended, discovery is exchanged, records arrive, depositions are taken, experts enter the case, and new evidence changes the significance of what came before.

That evolving intelligence layer feeds drafting. The motion is not being written next to the case file. It is being written from a connected understanding of the case file.

Motions from the record, firm priors, and jurisdiction rules

Generic drafting starts with a document type: "Draft a motion for summary judgment." Matter-based drafting starts somewhere else: What happened? What is established in the record? What remains disputed? Which testimony supports the position? What has the opposing party admitted? Which authorities govern the issue in this jurisdiction? How has this firm framed the argument before?

We bring those inputs together for motions such as summary judgment, motions in limine, motions to dismiss, and motions to compel.

  • The matter supplies the facts. Pleadings, records, discovery, depositions, correspondence, and other case materials establish what can actually be said about this dispute.

  • Firm priors supply institutional knowledge. Prior briefs can inform structure, argument organization, recurring language, citation conventions, and the way the firm presents an issue.

  • Jurisdiction rules constrain the work. The relevant procedural and substantive requirements affect what must be shown, how an argument should be framed, and what belongs in the filing.

Firm voice is more than tone. It appears in how a brief is structured, how the issue is introduced, which arguments come first, how record citations are used, and how much explanation a partner expects before the requested relief. We use a firm's prior work product to help drafts reflect those patterns. A prior brief should not become a factual template for the next case. The current matter still controls the facts, the current jurisdiction controls the law and procedure, and the legal team controls the position ultimately taken.

Discovery drafting has a consistency problem

Discovery creates a different risk. Each individual response may look reasonable while the matter as a whole becomes inconsistent. An interrogatory answer may characterize an event one way. A later response may describe the same event differently. An RFA may establish a fact that is overlooked when responding to another request. A pleading, verified response, deposition, or earlier admission may constrain what can safely be said later.

That is why discovery drafting needs memory across the matter. We can draft interrogatory, RFP, and RFA responses while cross-referencing the matter file, prior discovery in the same case, and admissions already in the record. The point is not simply to generate an answer faster. It is to make the existing litigation record visible while the next answer is being prepared.

Within Turbo Law, those materials connect through the matter's Fact Graph. A new record can change the understanding of an earlier event. A deposition can strengthen or contradict a factual position. An admission can become relevant to a motion being drafted months later. The drafting system should see those relationships because the lawyers reviewing the draft need to see them too.

Open the source, then propose, verify, decide

One of the most important questions in AI-assisted litigation drafting is simple: Where did that fact come from? A fluent sentence is not evidence. A plausible statement is not part of the record merely because a model generated it confidently.

We design key conclusions and factual propositions so they can be traced back to the underlying matter sources. Through the Fact Graph, teams can open the supporting page and line rather than treating the generated draft as its own authority. A motion may depend on allegations from the pleadings, an interrogatory answer, deposition testimony, an underlying record, and governing case law. Those sources should remain distinguishable even when the argument connects them. The Fact Graph provides the connective tissue. It does not erase the underlying sources.

ABA Formal Opinion 512 identifies duties implicated by lawyers' use of generative AI, including competence, confidentiality, supervision, candor, and communication. Florida Bar Ethics Opinion 24-1 emphasizes that lawyers remain responsible for their work product and professional judgment. NCSC guidance stresses human review of AI-generated content and risk-based oversight, particularly where AI is used for drafting, research, or decisions affecting legal rights.

Our operating model is intentionally straightforward: the platform proposes, staff verify, partners decide. Turbo Law can assemble matter intelligence, identify relevant connections, produce drafts, and place sources behind factual propositions. It does not convert generation into legal judgment.

Continuity across the matter is the real advantage

A motion should know what discovery established. A discovery response should know what the answer admitted. A mediation statement should know what the depositions changed. A new record should be capable of changing the team's understanding of an issue that first appeared in the complaint.

That continuity is difficult when drafting, research, document review, chronology work, and strategy occur in separate tools with separate context. We connect those workflows through the same matter intelligence layer. Pleadings, discovery, case law, records, case specifics, prior drafts, and subsequent developments become part of an evolving representation of the litigation rather than isolated inputs to separate prompts.

The question is not whether AI can write another legal document. It is whether the draft understands enough of the matter to deserve a lawyer's attention in the first place.

Related reading

Earlier pieces in this series:

Sources

Continuity across the matter is the real advantage

A motion should know what discovery established. A discovery response should know what the answer admitted. A mediation statement should know what the depositions changed. A new record should be capable of changing the team's understanding of an issue that first appeared in the complaint.

That continuity is difficult when drafting, research, document review, chronology work, and strategy occur in separate tools with separate context. We connect those workflows through the same matter intelligence layer. Pleadings, discovery, case law, records, case specifics, prior drafts, and subsequent developments become part of an evolving representation of the litigation rather than isolated inputs to separate prompts.

The question is not whether AI can write another legal document. It is whether the draft understands enough of the matter to deserve a lawyer's attention in the first place.

Related reading

Earlier pieces in this series:

Sources