BridgeScout is an agent for small consumer brands going abroad ("出海").
Tell it what your customers already love ("we're like Pop Mart + Sanrio"), and it uses Qloo's cultural taste graph to answer what a $20k market-research study would: which city first, which neighbourhood for the pop-up, which shops to pitch, which brands to co-brand with, which artists/podcasts sit next to your audience, and who that audience is. Every claim links to the exact Qloo request behind it (evidence E1…En).
Why Qloo is required: an LLM alone will say "launch in Los Angeles" for every Asian lifestyle brand. BridgeScout instead measures, per city, how strongly the local population over-indexes on your reference entities, then maps that affinity down to neighbourhood heatmap cells and cross-domain partners (places, brands, artists, podcasts). Swap the references and the answer changes — because taste changes.
Agent loop (tool-calling planner with scripted fallback):
- resolve_references → Qloo /search
- taste_dna → /v2/insights (tags)
- suggest_markets / score_markets → destinations + location affinity
- hotspots → heatmap
- local_partners → place | brand | artist | podcast
- audience → demographics
Live demo (Render): https://bridgescout.onrender.com Source: https://github.com/dongwei-polyred/bridgescout Landing / runbook: https://dongwei-polyred.github.io/bridgescout/
Log in or sign up for Devpost to join the conversation.