Common questions.

Jubal is a desktop AI partner for your firm. You point it at a folder of matter files and it builds a private index — full text plus structure plus cross-references. Then you can ask it questions across the matter, draft new documents in your firm's voice, surface inconsistencies in deposition transcripts, and check contracts against your playbook. Every answer cites the documents it came from. Each attorney runs their own index; for multi-attorney matters, Jubal supports secure firm-scoped sync. Only the small slices of text needed to answer a specific prompt are sent to the frontier model, behind a no-training agreement.

Mac and Windows at launch. Linux is on the roadmap but not committed for the founding cohort. If Linux is critical for your firm, tell us in the qualifier — it changes our priority.

The local index is fully local — search, retrieval, and matter-folder operations run on your machine without any network round-trip. Generative answers require a network connection because they call out to a frontier model API. We are working on offline drafting using local models for the firms that need it; for now, treat Jubal as a desktop app that occasionally calls a model API.

Yes. Each attorney runs Jubal on their own machine, with their own private index. For multi-attorney matters that span offices, Jubal supports secure firm-scoped sync — content stays encrypted in transit and at rest, and access is scoped to seats your firm authorizes. Firm-wide knowledge like your past briefs, clauses, and precedent can also live alongside your existing shared storage (network drive, OneDrive, Dropbox, iCloud, NetDocuments) when that's where your firm wants it. Your firm picks the model that fits.

Yes — a mobile companion ships at launch. It isn't the full product (deep work like full-matter analysis, drafting, and contract review against your playbook lives on the desktop), but it lets you stay in touch with your matters from anywhere: read incoming questions and updates, send quick prompts, review drafts. Partial access while you're away from the desk, with the same trust posture as the desktop.

Each attorney runs a private index on their own machine — no chunking into a vendor-hosted vector database, no upload of your raw documents to a Jubal cloud. For multi-attorney matters that span attorneys or offices, Jubal can sync your firm's content securely across the seats your firm authorizes; content is encrypted at rest and in transit, access is scoped to your firm, and we never use your matters to train any model. When you ask a question, only the small slices of text needed to answer that specific question are sent to the frontier model API, behind a no-training agreement. The /security page goes into the architecture in detail.

Yes — the architecture is built for it. Each attorney runs a local index on their own machine; for multi-attorney matters, Jubal supports secure firm-scoped sync with content encrypted in transit and at rest. Model providers run under contractual no-training posture. Audit logs and per-matter walls are first-class. We're happy to walk through your firm's security questionnaire — write to security@jubal.law and we'll respond within three business days.

We track the published ethics opinions across the major state bars and design Jubal to comply with the most restrictive of them. Local-first architecture, no-training model posture, citation-grounded outputs, and explicit human-in-the-loop drafting are direct responses to the ethics-opinion landscape. We are putting together a public roundup of the relevant opinions; ask us in the qualifier if you need it for your jurisdiction.

Most. The provisions that trip up other legal AI tools — no training on client data, no third-party hosting of privileged documents, encryption at rest and in transit, audit logs, named subprocessors — are all baseline for Jubal. Send us your firm's most-restrictive client OCG and we'll walk through it section by section. security@jubal.law.

Founding cohort opens summer 2026. We're inviting 100 firms first; the broader beta opens after the cohort has had a few weeks to put pressure on the product. Cohort firms get founding pricing locked in for the life of the contract.

Founding cohort pricing is available to firms on the waitlist. Public pricing publishes at general availability. We are not building a price-discrimination machine; the pricing will be transparent and per-seat with reasonable limits, not a custom-quote-only model.

Yes. Founding cohort firms lock in a discount for the duration of their contract — the rationale is straightforward, you took a chance on us before the product was finished. The exact percentage is set on a per-cohort basis and shared when we invite you.

We're optimizing for fit, not for prestige. The qualifier on the waitlist tells us your size, practice areas, current AI tools, and the work you'd most like an AI partner to take off your plate. We invite the firms whose use case our v1 best serves — meaning fewer disappointed customers and more useful product feedback. Submitting the qualifier moves your firm to the priority queue.

Jubal is built by a small team of software engineers and lawyers. Co-designed with a 40-year law firm we've worked with for three decades — they sit in our standups and use what we ship before anyone else does. The Jubal corp entity and the partner firm are separate businesses with a strict data wall between them; their value is craft and judgment, not access to other firms' work.

Our partner firm is BocaLaw.com, a 40-year practice in Boca Raton. Ken Ronan is our principal there and a co-designer of the product. We've worked with Ken and BocaLaw.com for thirty years across other projects; Jubal is the one we built together. The Jubal corp entity and BocaLaw.com are separate businesses with a strict data wall between them — their value to Jubal is craft and judgment, never access to other firms' work.

We will share funding details with founding cohort firms. The headline: Jubal is funded with the long-term commitment to serve solo and midsize firms — not to chase the AmLaw 100 with a different tool. The structure of the company reflects that focus.

United States. Engineering and partner-firm offices are both US-based. Customer data and customer-data-derived state stay in US infrastructure. International expansion is on the roadmap; for v1, we are an American product for American firms.