Arsalan Younus.
Back to Projects

Advisify

AI immigration assistant that delivers personalized pathway recommendations grounded in the user's CV and preferences, not generic advice lists.

The Business Problem

Immigration options change by country, visa category, and timing. Generic search and static articles do not personalize to someone's CV, goals, or the current policy landscape.

Applicants need guidance that respects their preferences and stays aligned with what is actually available, not one-size-fits-all advice.

The Technical Solution

I built structured intake combining conversational profile data with CV understanding, then layered retrieval and grounding on current program and rules material.

An LLM-driven experience explains pathways, trade-offs, and recommendations with clear boundaries around legal advice.

The Scalability Factor

Architected on AWS with containerized services. CI/CD pipelines via GitHub Actions automate deployment as product direction evolves.

Business Impact

Users receive preference-aware pathways and recommendations grounded in their documents and stated goals.

Clear narrative from profile to actionable next steps instead of scattered research.

Built with

LLMs
RAG
Python
LangChain
Document AI
AWS