ExpatHaven Hub.

Relocating to a new country is one of the most information-dense decisions a person makes — visa types, tax residency rules, cost of living, neighbourhood safety, healthcare access. The answers live across hundreds of disconnected government portals, Reddit threads, and Facebook groups.

Expats were spending weeks stitching together information from a dozen sources, none of which understood their personal situation. There was no platform that took your profile and gave you a personalised, actionable relocation plan.

We built Expat Haven Hub to fix that — an AI-powered assistant that turns a 10-minute quiz into a full country recommendation with visa guidance, tax breakdowns, living cost estimates, and a personalised news feed.

Sole engineer on the project alongside the team lead. Owned the assessment quiz flows, LLM recommendation engine, personalised news feed, country comparison modules, AI Q&A video features, and the admin panel — end to end.

Led the AI orchestration layer: prompt engineering for country matching, integrating OpenAI and Gemini APIs, and building the web scraping pipelines that populate country-specific data. Also handled OAuth 2.0 (Google Sign-In), Payload CMS integration, and production deployment.

Started with the data problem. Country information on visas, taxes, and cost of living doesn't come in a clean API. Built scraping pipelines to collect, normalise, and version data across 50+ countries.

The LLM recommendation layer required several prompt iterations. Early outputs were generic. Adding structured output schemas and injecting scraped country data as context dramatically improved match quality.

The admin panel was a significant engineering effort: user management, AI usage monitoring, per-user LLM cost tracking, Q&A video generation workflows, analytics dashboards, and saved items management — all with full RBAC.

A platform that gives any expat a personalised relocation roadmap in under 10 minutes.

The quiz feeds into an LLM pipeline that outputs country matches with expat profiles, key traits, and expert tips — grounded in real scraped data. Users get a persistent news feed tailored to their recommended countries, plus modules for visa details, tax comparisons, living cost estimation, and property insights.

An AI-powered Q&A video feature lets users ask relocation questions and receive video-based answers generated via OpenAI and Gemini — a capability no other platform in this space currently offers.

Platform handles end-to-end expat assessment and delivers personalised recommendations across 50+ countries with real-time scraped data and dual LLM support.