Your retail-savvy
AI & digital partner
Butik builds AI assistants for store teams, WeCom conversation analytics, AI sales coaching and a private livestream platform for retail brands. We stay with you from first conversation and POC through to rollout.
Brands we've worked with
Who we are
We know classic retail digital inside out, and we're excited about what AI lets every person do.
We don't talk about AI in the abstract. We start from real business situations and help companies move faster with AI, steadily. No hype, and no standing still.
Companies adopting AI are often held back by blind spots and by where their systems are today. We know retail, we know how businesses and their tech actually run, and we know AI. It takes all three to build AI that is genuinely useful.
Solutions
Retail AI Assistant
An AI assistant for client advisors. It helps them plan the day, catch sales opportunities and serve customers well.
Learn more →
DECISION
A WeCom conversation archive and analytics platform. It helps brands manage chat data, turn it into business insight, and feed it to AI agents.
Learn more →
Retail AI Coaching
Store training that combines practice with data. After each round we track business results and adjust what teams practise next, from customer role-play to product knowledge.
Learn more →
Butik Livestream
Six years of private livestreams for brands in China, built for high-value products and customers who expect an immersive experience.
Learn more →
Services
AI consulting & POC
From solution design, architecture and use-case selection to delivery and follow-up, we work alongside you the whole way.
Retail FDE partnership
Forward-deployed engineers (FDEs) who understand both AI and retail work with your team on the areas that matter to you.
Enterprise digital systems
Classic digital work: Mini Programs and e-commerce, WeCom, CRM and clienteling, loyalty programs and system integration.
Livestream operations
Studio hardware, on-site production, host training and day-to-day support.
Start with a POC
We usually suggest starting with a small POC: in two to four weeks you get a current-state review, a use-case shortlist and a working prototype. Then decide what's next based on results.
