User Experience Flow

How REX Works

From craving to reservation — seven intelligent steps

01

Natural Input

User describes what they're craving in plain language on the homepage. The query carries into the chat experience seamlessly.

Next.js sessionStorage
02

Context Extraction

AI parses mood, vibe, dietary needs, party size, price range, and occasion from the conversational input.

Gemini AI NLP
03

Follow-Up Questions

If key context is missing, Rex asks targeted follow-ups — refining the search before wasting your time with wrong results.

Conversational AI
04

Smart Matching

Google Places API serves as primary data source, with Supabase as intelligent fallback. 15+ variables matched against real-time restaurant data.

Google Places Supabase
05

Explained Results

Rex delivers recommendations with personality — each pick comes with a clear explanation of why it fits your specific ask.

Explainable AI
── optional loop ──
06

"More Options" Pagination

Not feeling it? Ask for more. Rex paginates through additional results with offset-based retrieval — no repeated suggestions.

Pagination Context-Aware Chips
07

Rex Gallery & Experiences

Every conversation and recommendation is persisted. Browse past searches by intent, revisit saved picks, and log dining experiences with ML-enhanced tags.

Supabase ML Tags