Sansiri sets a new standard for home buying and ownership in Thailand with Generative AI on AWS
Sansiri has rolled out Thailand's first large-scale Generative AI solutions in real estate, delivering faster customer responses (30%), higher satisfaction scores (25%), and smoother transactions across hundreds of projects. This isn't a pilot. It's a stack that touches day-to-day decisions for buyers, owners, and teams on the ground.
Built on Amazon Web Services, Sansiri's approach shows how AI can tighten service quality while reducing operational drag. The focus is clear: faster answers, fewer bottlenecks, and data that's accessible when people need it.
What buyers and owners get right now
At the center is ChatX Assistant, powered by Amazon Bedrock using Anthropic's Claude 3.5 Haiku and Claude 4 Sonnet. Think of it as a digital concierge that gives staff and buyers immediate answers on financing, zoning, construction updates, and project timelines.
Customers can check rules, options, and build progress across more than 500 projects in seconds. That helps decisions happen sooner and reduces repetitive workload for sales and service teams.
How operations are getting leaner
Sansiri's OCR system, also built on Bedrock, handles 50,000+ invoices per month with over 90% accuracy. Document processing time is down by half, which frees teams to focus on customer care and delivery.
The company is now exploring Agentic AI to add workflow automation on top of OCR-so tasks trigger automatically in line with enterprise processes and controls. Less manual follow-up, more consistent execution.
A stronger tech backbone for growth
On the engineering side, Amazon Q Developer has lifted software development productivity by 30% and cut compute costs by 20%. That keeps the platform scalable as demand rises and new projects come online.
Why this matters for real estate and construction leaders
The sector is moving fast on digital. Sansiri's results show how AI on AWS can improve customer trust, shorten response times, and bring clarity to complex decisions without adding headcount.
It's a practical playbook: centralize project data, expose it through an assistant, automate document-heavy workflows, and keep building velocity high with AI-native developer tools.
Practical takeaways you can apply
- Start with a concierge use case. Index policies, unit data, financing FAQs, timelines, and procurement rules so your teams and buyers get instant answers.
- Connect the assistant to source systems (ERP, CRM, DMS) for live updates on pricing, inventory, construction progress, and handover schedules.
- Treat OCR as a process, not a feature. Pair extraction with validation rules, approvals, and audit trails to reduce exceptions.
- Pilot Agentic AI for repetitive back-office tasks: invoice matching, drawdown checks, permit status alerts, and site issue routing.
- Measure what matters: response time, CSAT, error rates, cycle time to contract, and cost per transaction.
- Use AI coding tools to keep your backlog moving while controlling cloud spend.
What leaders at Sansiri and AWS are saying
Sansiri's technology leadership points to faster service, clearer information, and more personal support for buyers-plus better tools for employees to improve community living. AWS Thailand highlights the region's momentum and how services on AWS can streamline operations and raise the bar for the home buying and ownership experience.
Resources
- Amazon Bedrock for building and scaling GenAI applications.
- Anthropic Claude model overview and capabilities.
- Hands-on Claude certification to upskill teams on practical workflows.
Bottom line
Sansiri's deployment shows how AI moves the needle when it's tied to real workflows: faster decisions for buyers, fewer manual steps for staff, and a tech stack that scales. For developers and builders, the blueprint is here-and it's ready to be adapted to your portfolio.
Your membership also unlocks:
