From CV to a complete EB-1A package, run by an AI agent.
Upload your CV. The agent assesses your case, maps your evidence to all 10 USCIS criteria, plans your route, drafts only what your evidence supports, and audits it like a USCIS officer before you export. We stop where filing begins.
Most tools draft text. This one reads your record first.
End-to-end, not a toolbox
One workflow from the first CV read to a review-ready package — not five disconnected tools you have to stitch together.
Field-aware reading
Reads stat genetics differently from ML. Counts only independent citations, and knows which venues and awards actually carry weight in your field.
Grounded in your evidence
Every claim traces back to a paper, an award, or a line you wrote. The agent refuses to write what your record can't carry.
Built for researchers
Tuned for the postdoc record — papers, citations, reviewer panels, recommenders. A narrow door, on purpose.
Two ways in. Same workflow — the free tier just stops earlier.
Readiness call
Upload your CV. The agent runs the assessment and returns a verdict — Ready / Borderline / Not yet — with 3–5 specific next actions tailored to your record.
Full preparation
Unlocks the criteria evidence map, route planning, drafts of all six artifacts, and a pre-export USCIS-style audit.
After export, you take the package to a U.S. immigration attorney for review before filing. We're not a law firm — we don't pick the lawyer, you do.
Five stages, all grounded in evidence you actually have.
The free tier covers Stage 1. The $399 package covers all five.
- 01
Assess
FreeThe agent reads your CV and any materials you upload, end-to-end, and returns a verdict — Ready / Borderline / Not yet — plus the criteria your record may already support and the gaps you'd need to close.
- Independent citations only — self-cites and coauthors excluded before counting.
- Field-aware: knows which venues count and which awards are real in your field.
- 02
Map
PaidEvery piece of evidence you have, mapped to all 10 USCIS criteria. Not a one-time report — the live case state the agent keeps updating as you add material and revise.
- Each criterion shows MET / PARTIAL / MISSING with its supporting evidence underneath.
- Gaps are explicit — you see exactly what's missing before drafting begins.
- 03
Plan
PaidFrom the map, the agent recommends which 3–4 criteria to attack, which need more material, and which aren't worth chasing. Not "you might qualify" — "here's your route."
- Field-aware: a year-3 stat genetics postdoc gets a different route than a CS industry researcher.
- Lists exactly what to ask each recommender to cover.
- 04
Draft
PaidEvery artifact you need, grounded in your evidence — six documents you can hand to a lawyer or use yourself:
Personal StatementPetition letterRecommender templatesExhibit listSubmission checklistRisk audit- Every sentence links back to the paper, award, or material it came from.
- Ask for a claim your evidence can't carry and the agent pushes back, narrows it, or refuses.
- 05
Audit & Export
PaidBefore you export, the agent switches sides and reads your final draft like a USCIS officer trying to RFE it. Weak claims, missing exhibits, contradictions — surfaced before you ship. Then export to PDF or DOCX.
- Every revision is a version — roll back, compare, keep both.
- The case waits for you. Close the tab, come back in three weeks, pick up where you left off.
An evidence map you can read like a referee report.
For each of the 10 USCIS criteria, the map shows what evidence supports it, what's missing, and where a claim outruns its evidence. Every line links back to a paper, an award, or a paragraph you wrote.
- i.MetAwards for excellenceASHG Best Paper (2024) · 2 departmental honors
- ii.PartialMembership requiring achievementASHG member; verifying admission criteria
- iii.MetPublished material about youNature News commentary (2024) on PNAS paper
- iv.PartialJudging the work of others1 reviewer panel mentioned — needs documentation
- v.MetOriginal contributions of major significancePNAS 2024 (47 independent citations) · 3 first-author works
- vi.N/AAuthorship of scholarly articlesCovered by criterion v — not a separate claim
- vii.N/ADisplay at artistic exhibitionsNot applicable to this field
- viii.PartialLeading or critical roleLab lead on 2 funded subprojects — needs PI letter
The strongest agent isn't the one that writes the most.
The valuable part is restraint: reading the case, refusing unsupported claims, and drafting only what the evidence can carry. When the evidence is thin, the agent asks for proof, lowers the strength, or leaves the claim out.
“Make this contribution sound field-changing.”
Do independent citations, adoption examples, or third-party letters actually support that strength?
Use a supported, narrower claim — or pause drafting until the missing evidence is uploaded.
- Never trained onYour CV, drafts, and recommender info are used only for your case.
- EncryptedAll case material is encrypted at rest and in transit.
- Delete anytimeRemove your case and we remove the data.
- Not a law firmBring your own U.S. immigration attorney to file.
- EB-1A onlyWe don't do other visa categories.
Questions postdocs ask before they start.
Is VisaCanvas a law firm?
No. We don't provide legal advice, don't represent you before USCIS, and don't partner with attorneys. The product only does material preparation.
Do you have a partner attorney I can use?
No. You find a U.S. immigration attorney yourself for review and filing. That's an intentional boundary — we focus on material prep, not lawyer referrals.
How is this different from using ChatGPT to draft my petition?
Three things. (1) The agent reads field-aware — what counts in stat genetics isn't what counts in ML. (2) Your case persists across sessions: close the tab, come back in three weeks, the evidence map is still there. (3) It refuses to write claims your evidence can't carry, and every sentence traces back to a paper or material you uploaded.
Will my data be used to train AI?
No. Your CV, drafts, recommender info, and case data are used only for your case. They are never used to train models.
What's your approval rate?
We don't track or report approval rates. USCIS adjudication depends on factors beyond material quality, and any percentage would mislead you. We focus on whether your evidence map and drafts hold up against the criteria — not on outcome predictions.
I'm a year-1 postdoc. Should I even try?
Run the free readiness call. If the agent says “Not yet,” it tells you the specific evidence to accumulate and roughly when to come back. The free tier exists for exactly this decision.
What if I'm in industry, not a postdoc?
You can still use it, but the agent is currently optimized for the researcher record — papers, citations, reviewer panels, academic recommenders. Industry-only records will work less well right now.
Can I just file the package myself after export?
We don't recommend it. We strongly recommend a U.S. immigration attorney review the package before you file. We can't introduce you to one, but the export is in a format your attorney can review directly.
Start with the free readiness call.
Upload your CV. Get a verdict in two minutes. If we say “not yet,” we tell you why and what to fix. If we say “ready,” you decide whether to unlock the $399 package.
Start your session →VisaCanvas is not a law firm and does not provide legal advice. We recommend having a U.S. immigration attorney review your package before filing.