Carenavio
Telemedicine Platform
Full Stack2026
VeriPass
Legal Tech Platform
Founders spend thousands on immigration lawyers just to learn they don't qualify for their target visa. I built a platform that models US immigration law into scoring algorithms, assesses eligibility in minutes, and recommends optimal visa pathways. The system achieves 98% accuracy and has become a qualified lead funnel for the company's legal services.
Company
Thelix
Role
Full Stack Engineer
Duration
6 months
Year
2026
Tech Stack
Next.jsNode.jsPostgreSQLTailwind CSSOpenAI
The Problem
Founders waste months and thousands of dollars pursuing wrong visa categories. Initial legal consultations cost $500+ just to determine basic eligibility.
The Solution
Built a rules engine that encodes O-1A, EB-1A, and EB-2 NIW requirements into weighted scoring algorithms. Users answer structured questions, and the system predicts eligibility with reasoning the user can verify.
Technical Challenges
1
Translating ambiguous immigration law into deterministic scoring logic
2
Designing UX that guides users through complex legal questions without overwhelming them
3
Building confidence in AI recommendations for life-changing decisions
4
Creating conversion funnels that qualified leads without feeling predatory
Results & Impact
500+
Users
founders assessed
35%
Conversion
to paid consultations
98%
Accuracy
eligibility predictions
Lessons Learned
Key Takeaways
- Legal tech requires deep domain understanding—you can't abstract away the complexity
- High-stakes UX needs to show reasoning, not just results
- The right assessment tool can qualify leads better than sales teams
