How Axrail Helped Castlery Launch Casa, a Personal AI Shopping Assistant
How Axrail built Casa, Castlery's AI shopping assistant on Amazon Bedrock, cutting response times, personalizing product discovery, and freeing staff from manual lookups.
Instant
responses to product questions and order-status enquiries, day or night
AWS Premier Partner with GenAI expertise.
See how Castlery redefined customer experience with Axrail's generative AI
Castlery's digital-first furniture brand put its online shopping experience through Casa, an Axrail-built AI Frontline Agent on Amazon Bedrock, answering product and order questions instantly so shoppers get the same personal attention as in-store.
About Castlery
Castlery is a Singapore-founded, digital-first furniture brand that designs, manufactures, and retails high-end home furniture, delivering to customers across Singapore, Australia, the UK, and the USA. Its design-led approach — combining award-winning in-house aesthetics with durability and quality — has made Castlery a preferred choice for customers furnishing their homes online.
The challenges
Slow first response to online queries
Online shoppers with many brand options at their fingertips expect immediate attention. Without it, customers navigating product questions were more likely to hesitate or abandon a purchase before completing it.
Manual lookups pulling staff from higher-value work
Every incoming product query that required a staff member to manually look up information was time not spent on other parts of the business — a workload that doesn't scale cleanly with online traffic.
Fragmented visibility into buyer trends
Without a systematic way to capture how customers engage across touchpoints, Castlery's team had limited data to inform how the buying journey could be accelerated and personalized further.
The solution
To bring the immediacy of in-store service online, Axrail built Casa on Amazon Bedrock: a generative AI Frontline Agent that engages Castlery’s shoppers directly on the platform, from the first query through to purchase support.
Casa answers customer queries the moment they’re asked, cutting the wait shoppers would otherwise spend on a human agent, and helps them find what they love faster with personalized recommendations rather than generic search results — echoing the guided experience of an in-store sales assistant.
Customers can track order status and get instant answers about past purchases without waiting on a support queue. And when a query gets complex, Casa hands off to a human agent with no friction and no repeats, so the customer never has to start their question over from scratch.
Casa runs on Amazon Bedrock, which gives it natural language conversational capabilities that recognize the sentiment and nuance of each incoming query, keeping responses relevant and comprehensive — and setting up a foundation Castlery can extend toward agentic and multi-agent workloads as the next step in its digital-first strategy.
See Casa in action
A recreation of Castlery's own assistant, built off the real interface — try one of the prompts a shopper would actually type.
An interactive recreation of Castlery's own Casa interface, built to show how it actually responds — not a screenshot.
Conclusion
Axrail's Casa brings Castlery's online shoppers the same immediate, personalized attention as an in-store sales assistant — answering questions the moment they're asked, guiding product discovery, and freeing staff from manual lookups for higher-value work.
Axrail is an AWS Premier Tier Partner delivering Applied AI and Digital Workforce solutions across Singapore and Malaysia, with Generative AI and Retail Services AWS competencies. Castlery selected Axrail through AWS's partner network on the strength of that track record helping enterprises deploy their first AI workload — with a methodology that starts from the measurable business outcome, not the technology, and a clear line from the AI shopping assistant to both Castlery's top-line and bottom-line.
Instantresponses to product questions and order-status enquiries, day or night
More customer success stories
More results from teams who automated their document processes with Axrail.
How we transformed Sepang International Circuit's fan experience with generative AISeconds to answer enquiries that used to wait up to two hours by email
How we helped QL Maxincome improve CRM performance and cut costs by 90%90% lower running cost after retiring on-prem SQL Server licensing
How Axrail Helped FamilyMart Malaysia Automate Invoice Processing with Finance Operations AI2K-3K invoices processed every month, extracted and converted into POs automatically
How we solved Pelangi's manual sales processing with AI40% less time spent processing each purchase order
Get the same results for your team
Book a discovery call and we'll scope a focused first project for your team, running in your own AWS account.