Roamo
Personal project
Data & Software
2,026
01/
Overview
Roamo is a mobile app that helps users save, organize, and explore places they love - restaurants, bars, museums, attractions, etc. Users can add spots via Google Maps links, websites or Pinterest pins and the system intelligently extracts and enriches information like addresses, categories, images, and opening hours. Spots are grouped by canonical metro city while preserving neighborhood context, making discovery intuitive and globally consistent.
02/
Process
I conducted user research and interviewed users about their experience saving spots, identifying pain points with their current methods (e.g., fragmented organization, difficulty tracking visited spots, note taking). Based on this, I designed a multi-step workflow:
Quick Add: Users paste a link, which the system validates and routes to the correct extractor.
Data Extraction: Extractors fetch structured data from Google Places, websites (JSON-LD/OpenGraph), or social media.
AI Enrichment: Low-confidence or messy data is cleaned and enriched using AI to infer categories, summaries, or other missing metadata.
Confirmation & Storage: Spots with high confidence are auto-saved, while medium/low confidence items prompt users for verification. Metro city and neighborhood/admin context are stored to ensure global consistency.
03/
Key features
Quick Add: Auto-draft spots from links, with AI-assisted category and metadata enrichment.
Canonical City System: Metro city grouping with optional neighborhood/admin labels, preventing fragmented results.
Opening Hours & Suggested Time: Structured hours from Google Places; safe suggested visiting times if missing.
Interactive Map & List Views: Zoomable maps with dynamic markers and clusters at city/region levels.
Conflict Detection: Flags discrepancies between user-entered data and authoritative sources, allowing review.
Global Support: Reliable worldwide handling for cities, neighborhoods, and administrative regions.