A faster storefront, a mobile app and an AI support assistant for a D2C brand
Lumora needed to hold up through peak season and answer customers faster. We rebuilt checkout, launched apps and added an AI assistant.
- Checkout conversion after the rebuild
- +22%
- Of support questions answered by the AI assistant
- 38%
- To launch the AI assistant
- 3 weeks
- Client
- Lumora
- Industry
- E-commerce
- Timeline
- 2025
- Platforms
- Web store, iOS, Android, support tool
- Services
- Web platform, Mobile apps, AI solutions, Performance
The challenge
Lumora's store slowed down during every sale, checkout dropped orders under load, and the support team spent most of its time answering the same delivery and returns questions.
What we built
- A rebuilt storefront and checkout with server-side rendering, caching and a simpler three-step checkout.
- Mobile apps for iOS and Android launched in the same week, sharing one product catalogue and cart.
- An AI support assistant that answers from Lumora's own help centre and order data, and hands over to a person when it is not sure.
How we worked
We load-tested checkout at three times the previous peak before the sale season. The AI assistant was rolled out to 10% of conversations first, measured against the human team, then expanded.
They added an AI assistant to our support tool in three weeks, and it actually reads our docs correctly.
Technology
- Next.js
- Node.js
- Flutter
- PostgreSQL
- Redis
- OpenAI
- Claude
Planning something similar?
Tell us what you want to build, improve or automate. We reply within one working day.