For immigration attorneys
AI that reasons from the law.
The judgment stays yours.
Papers arms your firm with AI that reasons from the legal standard, not from sample filings — so you prepare cases like an AI-native practice while your attorneys keep the judgment, the clients, and the brand.
Dr. Ananya Mehta
EB-2 NIW · Computational biology
Case readiness
Included in handoff
Evaluation summary
Three prongs assessed
Evidence workspace
46 exhibits, grouped
Draft petition
8 pages, cited
Questions for counsel
2 open
AI reasoned from the standard. Your firm applies the judgment.
Open caseWe arm your attorneys. We don't replace them.
Immigration is moving to AI. Papers lets your firm move with it — not by handing your practice to an AI service, but by arming your attorneys with AI that prepares the work and leaves the judgment where it belongs.
Your firm keeps the client
We arm attorneys; we do not replace them. Your firm owns the engagement, the communication, and the continuing relationship.
Judgment stays with counsel
AI can evaluate, organize, and draft. The legal conclusions and filing strategy stay with your attorneys.
Sharing is consent-based
Clients choose which materials to share with a partner attorney, for a specific review. No automatic handoff.
Why our AI
We teach the AI how to think, not what to copy.
Most legal AI is a wrapper. It learns what good filings look like — from samples, style guides, and retrieval over the law — and produces output that looks right. Papers reasons from the standard the work has to meet.
Reasons like an attorney
Issue by issue against the controlling authority — the statute, the regulations, the agency's own policy, and the decisions that bind it — not paragraph by paragraph against a template.
Reasons like an adjudicator
Tests the record against the burden of proof and asks what a neutral adjudicator would conclude. Adjudicator-aware, not advocate-only.
Trained to distill legal reasoning
Built to learn how the law is actually applied, then tag every claim Documented, Inferred, or Unsupported.
Most legal AI — a wrapper
Trained on sample filings
Prompted to match a style
Retrieval over the statutes and rules
Teaches the AI what to do. The output looks right.
Papers — engineered for legal reasoning
Reason from the controlling authority, issue by issue
Think like the attorney building the case
Think like the adjudicator testing the record
Tag every claim: Documented, Inferred, or Unsupported
Teaches the AI how to think. The output is right because the reasoning is right.
Don't take our word for it. Read the reasoning.
Our evaluation method is open source. Inspect exactly how the AI reasons, or have your own attorneys and paralegals critique it.
Where judgment lives
The subtle errors are the dangerous ones. Your attorneys catch them.
When AI drafts and a professional reviews, the professional spots what the model missed and corrects it. A self-petitioner cannot run that loop — they can only trust the output. So the judgment belongs with counsel, and Papers is built to put your firm exactly there.
AI preparation
Handoff — the client chooses what to share, at any stage
Attorney judgment
AI prepares. Your attorneys are the harness no model replaces.
The evaluation funnel
Free evaluation is the front door. You get the prepared client.
Papers' public evaluation reaches candidates months before they would call a firm. They keep building in the workspace, and arrive as a structured, consent-shared case — not a cold inquiry.
Traditional evaluation
- 1Applicant submits a resume
- 2Attorney reviews for viability
- 3Decision: take, wait, or pass
Applicants who aren't ready to retain have nowhere to go. Most never come back.
With Papers
- 1Applicant starts the evaluation in Papers
- 2The workspace builds over weeks and months
- 3Evidence, drafts, and questions accumulate
- Your firm receives a structured record
Applicants who weren't ready at first arrive as clients with months of structured preparation behind them.
Practice growth
Make your firm busier, not just faster.
Other tools help you draft faster on the cases you already have. Papers also brings you cases — and meets clients at the moments they usually handle alone, delay, or never bring to counsel.
Exploring
2Wei Chen
Sofía Almeida
Building
2Daniel Okafor
Priya Nair
Pre-filing review
1Dr. Ananya Mehta
Petition review
1Jae-won Kim
RFE / NOID
1Dr. Rafael Ortiz
Immigration is flat-fee work — attorney time is the only variable cost. Every hour Papers moves off your attorneys is margin you keep, or a case you finally have the capacity to take.
What your firm receives
A prepared record, not a loose referral.
Not a directory and not a referral marketplace. Papers gives you a consent-based packet with the client's case map, evidence, and draft materials in one place.
Evaluation summary
Pathway fit, the proposed endeavor, documented evidence, assumptions, and the areas needing attorney review.
Case theory · three prongs assessed · flagged items noted
Evidence workspace
Evidence organized so your review starts from structure instead of a folder of scattered documents.
Exhibits grouped by claim, source, and evidentiary role
Draft materials
Petition draft, recommendation letter drafts, exhibit list, response planning, and a submission checklist.
Petition draft · recommendation letters · exhibit list
Attorney review context
The client-selected materials, plus the specific questions and decisions the client wants reviewed.
Applicant questions · consent recorded
Two ways to partner
Start with prepared clients. Grow into an AI-native practice.
Two ways to work with Papers: receive clients who arrive already prepared, or run the whole AI-native workflow under your own brand. We're onboarding a small group on each.
Live today
Assist by Attorney
Receive applicants who prepared their case in Papers and now want a second legal review or representation.
Design partner track
Powered by Papers
Run AI-native evaluation, intake, and case preparation under your own brand, review process, and client relationship.
Powered by Papers
Your firm, your clients, our AI.
Become an AI-native practice without becoming a tech company — or surrendering your clients to one. White-labeled evaluation, intake, and case preparation under your brand; your attorneys review, comment, and finalize.
AI preparation notes inline
Every passage carries its evidentiary basis, so review starts from the strongest version of the record.
Attorney comments and control
Your team comments, requests evidence, and decides what gets finalized. Nothing files itself.
Controlled visibility
Clients share specific materials, for specific reviews, with consent recorded.
Document
Exhibits
§ 3. Well-positioned to advance the endeavor
The petitioner has demonstrated advanced expertise in interpretable machine learning for biomarker discovery, with a record of adoption extending beyond her immediate collaborators.
Two independent research groups at Stanford Medical Center and NIH NCI have incorporated her published methodology into their own active studies (Exhibits 12, 13).
The record also includes a documented NIH R01 grant award and citation analyses showing above-field-average uptake of her primary methods paper…
AI preparation note
Well-documented passage. Exhibits 12 and 13 both confirm independent adoption. The Prong 3 adoption gap is flagged separately.
Attorney comment
Request one additional user letter before finalizing this prong. Two adoptions are strong; a third independent source removes the risk at NOID stage.
Confidentiality & trust
Your client's data stays your client's data.
Recent federal rulings have made it concrete: casual use of public AI tools can forfeit confidentiality — and the privilege that depends on it. Papers is built the other way: enterprise terms, no training on your data, and work that happens inside your firm's control.
Not the consumer tools
Papers runs on enterprise AI under contract — not the public apps whose terms permit training on inputs and disclosure to third parties.
Never used to train a model
Client information, and the work product built from it, is processed under enterprise data terms — not used to improve any model, and held inside Papers' own cloud.
Inside your firm's control
Preparation happens in your firm's workspace, under your attorneys' supervision — the posture courts look to, not an unsupervised public chatbot.
And the lines we don't cross
Papers is not a law firm and does not advise on privilege; your firm applies its own professional judgment. Data-processing terms and security details are available on request.
Become a partner
Partner with Papers.
We're onboarding immigration attorneys and firms who want prepared clients today — and a voice in the AI-native workflow we build next.