Sparkl

Sparkl is a two-sided marketplace connecting vacation-rental hosts with cleaning professionals. Hosts post cleaning jobs tied to guest turnovers; cleaners browse nearby jobs and claim them. It was built for the Meta Global AI Developer Hackathon. Core MVP flow: a host signs up, adds a property, and posts a cleaning job (via form or a chat assistant with price hints and voice input); the job appears on an open job board where cleaners browse ranked listings; either side can initiate (cleaner requests, or host offers to a suggested cleaner), and the job is claimed once both sides agree; host and cleaner then coordinate via per-job chat through the job lifecycle (open → claimed → in_progress → completed); afterward both leave mutual reviews. Key differentiator: ranked, bidirectional matching — hosts see suggested cleaners, cleaners see jobs ranked by fit (rating, distance, availability, completed jobs, acceptance rate, response time), with deterministic match reasons shown in the UI, not just a newest-first list. Built English-first with i18n scaffolding for future Spanish support; no in-app payments, auto-assignment, or admin panel in MVP scope.