Skip to Main Content

Asoro Technology

Built at Meta Global AI Developer Hackathon · Oct 26, 2026 · Remote

Asoro Technology — Demo video

RecyGo turns one crowded photo of used items into a catalog-ready collection. Point the camera at a shelf or a sack of bottles; RecyGo separates the objects, cuts each one out, fills in the form, then offers what to do with it: sell it, give it away, or Fuse it with other items into something new. The problem is that listing beats discarding only if it is cheap. One listing means eight fields for an object worth a few thousand rupiah, so reusable material gets thrown out. RecyGo compresses that to one shutter press plus a review screen. Muse Spark reads the photos and returns strict JSON, whitelisted and enum-clamped server-side so a hallucinated category cannot corrupt the database. SAM 3.1 segments the same photo into individual instances from text prompts; the app crops and masks each bbox into a transparent PNG cutout on a background isolate, and a bulk review screen merges duplicates and saves everything at once. Muse Image illustrates each Fuse idea, using the user's own photo as an edit reference. Downstream is real: Supabase Postgres with row-level security, Community listings (For Sale / Free / Fuse Together) with search, distance sort and WhatsApp handoff, and a Waste Bank directory on OpenStreetMap. Flutter on Android, iOS and web; a single Rust service holds the Meta credentials; demo mode runs offline without keys.

Team