This post walks through what that looks like in practice. If you're new to either tool, the next section is a quick orientation. If you're already familiar with both, skip to the workflows.

A short primer

Paximal is an agentic AI platform for patent application drafting. Through an alignment process — a structured back-and-forth that replaces prompt engineering — Paximal generates complete, fully-enabled drafts in hours rather than weeks. With the new MCP, those agents are invokable directly inside Claude Cowork.

Juristat is the leading patent intelligence and analytics platform. It aggregates and structures the public USPTO record: examiner reports, art unit statistics, allowance rates, office action histories, appeal outcomes, claim-amendment patterns. Juristat's MCP exposes that data as queries — by examiner, by art unit, by application, by tag.

What changes when both are in the same conversation: the drafting agents are no longer operating from the surface layer of public filings alone. They can read the evidence of how the relevant slice of the USPTO actually behaves and draft accordingly. That's the lever the rest of this post is about.

Workflow 1: Art-unit-aware claim strategy, before you draft

A new disclosure comes in. Based on the technology, you have a reasonable guess about which art unit will examine the eventual application — say, one of the AI/ML art units inside class 706, or a fintech-adjacent unit in 705. In the old workflow, that guess didn't really influence drafting. You wrote the strongest claims you could and hoped the assigned art unit would agree.

With both MCPs in the conversation, the workflow runs differently. Before alignment, you ask Juristat for the relevant art unit's report: average number of office actions to allowance, allowance rate, the dominant rejection grounds, common claim-amendment patterns that survive. That snapshot lands in the conversation. You then start a Paximal project on the disclosure, with the art unit profile available as context. Paximal's drafting agents can take that profile into account — for example, by structuring the independent claims to anticipate the rejection grounds most common in that unit, by including dependent claims that mirror amendments shown to survive in similar prosecutions, and by writing spec support for fallback positions the data suggests you'll need.

The practitioner is still making the strategic decisions. What Juristat does is turn an instinct ("this art unit tends to bounce claims on § 101 grounds") into a specific, current, queryable fact. What Paximal does is fold that fact into a draft instead of leaving it as a mental note for later. The composite is a first draft that arrives already aware of where it's going.

Workflow 2: Examiner-aware drafting for continuations-in-part

For a brand-new application, you can't tell which examiner will pick it up — assignment is opaque from the applicant's side. CIPs are the case where examiner-aware drafting actually earns its keep. A continuation-in-part adds new matter to a parent that's already been examined, typically lands with the examiner who handled the parent, and — unlike a continuation or divisional drawing claims from the existing disclosure — involves substantive new drafting work where claim language and structure are genuinely up for grabs.

With Juristat connected, you query that examiner in both directions. On the success side: the claim-language patterns in their allowed cases, the shape of the dependent claim ladders that have moved efficiently through their docket, the kinds of spec support they've cited favorably. On the failure side: the rejection grounds they've raised most often, the claim structures they've consistently knocked down, the amendment language they've forced applicants into, the § 112 questions that recur in their docket. The full picture — what works and what doesn't — becomes input to alignment with Paximal.

The drafting agents can then shape the CIP's new claims toward language and structures this examiner has allowed before, build the dependent claim ladder around fallback positions the examiner has previously accepted, and write spec passages that head off the rejection grounds this examiner tends to raise. The judgment work — should we accommodate this examiner's tendencies, or draft to push past them? — stays with the practitioner. What changes is that the choice is now an informed one, made up front, against specific evidence about a specific individual who has already engaged with this family.

Workflow 3: Portfolio-calibrated drafting playbooks

The first two workflows are matter-level. The third is what the combination unlocks for in-house teams and the firms that serve them: drafting playbooks calibrated against your own portfolio.

In a typical setup, an in-house IP team has hundreds or thousands of applications in flight, drafted over years against evolving practice. Some are getting through cleanly; some aren't. The pattern that matters — "our drafting in this technology area produces more office actions than the median for this art unit" — is the kind of thing that's true for years before anyone notices. And even when it's noticed, knowing what to change about the drafting requires comparing your patterns against what the broader USPTO data shows is working.

With both MCPs in a recurring Cowork workflow, that comparison becomes routine. Pull portfolio-level outcome data, line it up against Juristat baselines for the relevant art units, and surface the technology areas where the gap is meaningful. From there, Paximal generates a revised drafting playbook for that area — specific guidance on independent claim structure, dependent claim composition, spec emphasis, embodiment selection — calibrated to what the data shows is actually working in that part of the USPTO. New applications in that area get drafted against the updated playbook.

For firm leaders and in-house IP teams, this changes a structural problem. Drafting strategy used to be embedded in the heads of senior partners who'd seen enough applications go through enough art units to have developed a feel for it. That expertise was hard to transfer and slow to update. Now it becomes a continuously refreshed asset — a queryable, updateable layer that any practitioner working on a new application can consult.

What this changes about practice

Three things, in order of how soon they'll show up.

First, the baseline quality of a first draft goes up, because the draft is now informed by examiner and art unit evidence rather than written in a vacuum. Junior practitioners benefit most here, but so does anyone working in a corner of the USPTO they don't see often.

Second, the first-pass posture of a new application improves. Drafts shaped against the actual patterns of the likely art unit — and, in CIP work, against the known examiner — land in a better position from the start: fewer iterations to get to a defensible scope. The compounding effect across a portfolio is meaningful.

Third — and this is the slower-burn change — who can do this work well expands. The senior-partner instinct about how a particular examiner reads claims, or how a particular art unit handles § 112, used to be the differentiator between a strong drafting practice and a weak one. It still is. But the instinct becomes legible. It becomes something an early-career attorney can consult, learn from, and check their work against. Over time, that levels the playing field within firms and between them.

We've said since the start that Paximal exists to amplify attorneys, not replace them. The composability with Juristat strengthens that thesis: drafting is now informed by the same underlying evidence base that shapes how applications actually fare at the USPTO, available to the same practitioner in the same conversation. The decisions are still the practitioner's. The information available when making them is dramatically better.

If you're a current Paximal customer, connecting both MCPs in Claude Cowork is a setup step we'll walk you through. If you're not yet, schedule a Paximal demo, a Juristat demo, or a combined demo of the two together — we'll show you the workflow in action.

The collaboration between Paximal and Juristat was, at the time we announced it, a strategic intent. What the MCP makes possible is the everyday version of it — the part that shows up in your Tuesday.