Skip to content
All work

Case study

AI Immigration Pathway Assistant

In plain English

Applicants see ranked immigration pathways from their profile before they commit to a country or a lawyer.

Free eligibility snapshot, paid deep-dive report, and an operator-managed catalog.

Who it is for

  • Skilled-migration applicants exploring where they stand before they hire a lawyer or commit to a country.
  • Immigration consultancies that need a self-serve assessment funnel: free eligibility snapshot, paid deep-dive, and an operator-managed catalog.

Problem

Most people pick a destination first, then discover scoring rules, occupation lists, and questionnaires after they have already committed time.

Generic articles do not match a person's CV or goals. Applicants need a guided, evidence-based view of eligibility, not a law-firm case file.

Solution

We reversed the order: destination, visa intent, and personal profile become ranked pathways, then a paid country-specific questionnaire and final assessment.

A Next.js applicant and admin app talks to a FastAPI backend. Deterministic scoring engines (Canada, Australia, New Zealand, USA eligibility) sit beside LLM-structured ranking and CV parse. Operators manage the visa catalog with human review before updates go live. The product does not file applications or present itself as legal advice.

Outcomes

A full product surface: marketing site, applicant wizard, AI-assisted assessments, Stripe-gated premium reports, and an operator admin on one stack.

Free eligibility snapshot, paid deep-dive and report: catalog as a product, not a silent scrape-to-live pipeline.

Next.jsFastAPIPythonMongoDBOpenAIStripeNextAuthDockerAWS ECR

How it works

Applicant web app, assessment engine, and immigration knowledge base with trusted services.
Applicant web app, assessment engine, and immigration knowledge base with trusted services.

From profile to pathway

Journey from profile through scoring and ranked pathways to a paid deep-dive report.
Journey from profile through scoring and ranked pathways to a paid deep-dive report.

What we built

Applicant journey

  • Profile wizard

    Destination, visa intent, personal profile, and family. Optional CV upload that fills profile fields.

  • Initial assessment

    Free eligibility snapshot: score, core pillars, and discoveries for chosen destinations.

  • Visa suggestions

    Ranked pathways, what-if modifiers, confirm a pathway, and a paid unlock.

  • Deep-dive and reports

    Paid questionnaire for the confirmed pathway, points breakdown, and a detailed assessment dashboard.

Operator admin

  • Dashboard and users

    Signups, completions, conversions, and revenue, plus user management.

  • Immigration catalog

    Countries, visas, and questionnaires in one hub.

  • Ingestion review

    Queue extract jobs, review diffs, and approve or reject catalog updates before they go live.

How it works

The Next.js app calls FastAPI. NextAuth holds the backend JWT; protected routes require a signed-in user, and admin requires the operator role. Profile, report, and payment routes are owner-scoped.

Scoring is deterministic per country program; the LLM is used for CV parse, visa-suggestion copy, and optional catalog enrichment, not as a black-box visa oracle. Catalog ingestion is review-then-publish.

Integrations

  • Stripe

    Checkout unlocks the paid deep-dive and report on the profile.

  • OpenAI

    CV-to-profile extraction, visa-suggestion structuring, and optional ingestion enrichment.

  • Google OAuth

    Applicant sign-in alongside email and password.

  • Country-source ingestion

    Operators refresh official immigration content with human review before it goes live.

Have a similar challenge?

Start a project