Advisify

Client Challenge

Immigration eligibility rules live scattered across government sites and PDFs, so most applicants only discover them after they've already committed to a country - with no easy way to compare destinations side by side. Immigration consultancies and agencies lose advisor time on manual intake for unqualified leads, with no self-serve funnel to filter and monetize early interest. Meanwhile, a catalog of hard-coded rules goes stale the moment a country updates its points table.

Objectives

• Give applicants a guided, evidence-based view of eligibility across multiple countries before they commit to one.
• Score eligibility deterministically per country program, keeping the LLM scoped to copy, CV extraction, and ingestion structuring - never the scoring decision itself.
• Keep the underlying visa rules catalog current through a reviewed ingestion pipeline, not a one-time hard-coded ruleset.

Solution Design

We built a Next.js applicant web app on top of a FastAPI assessment engine, with an immigration knowledge base in MongoDB kept fresh by operators.

Applicants build a profile through a guided wizard or CV auto-fill, then deterministic points engines - Canada's CRS and Federal Skilled Worker, Australia, New Zealand, and the USA - score their fit against every program. A free eligibility snapshot ranks the results; confirming a pathway and unlocking the paid deep-dive report runs through Stripe Checkout, with is_paid gating the Deep-Dive, Final Assessment, and Reports from a single flag. Every sensitive route is owner-scoped, and the browser never talks to MongoDB or Stripe directly - every call is proxied through Next.js to FastAPI. Behind the scenes, a polling ingestion worker discovers, crawls, and extracts rule updates from official sources; nothing publishes to the live catalog until an admin reviews the diff.

Implementation Process

Results & Business Outcomes