Thai startup’s AI assistant supports 20+ regional languages with up to 95% accuracy, helps organizations drive customer service improvements across Southeast Asia

Amazon Web Services (AWS), an Amazon.com company, announced that BOTNOI, a Thai generative artificial intelligence (generative AI) startup specializing in conversational AI virtual assistants, has built text-to-speech and speech-to-text AI platform BOTNOI Voice on AWS. Tailored to Southeast Asian languages and cultural norms using consented language data, BOTNOI Voice is enables local organizations to improve customer service with generative AI.

BOTNOI Voice supports more than 200 near-human voices in over 20 languages, with a focus on Thai, Indonesian, Vietnamese, Lao, Burmese, and Cambodian.

AI Tech Insights: AI-Driven Testing Solution Launchable by CloudBees Joins AWS Marketplace

BOTNOI has deployed its virtual assistant platform in more than 70 enterprises in Thailand such as Central Pattana (CPN), Bangkok Hospital, and over 10 million end-users have interacted with its AI assistants.

BOTNOI Voice users have developed a range of assistants, including AI nurses for patient pre-screening and pre-registration to help healthcare professionals classify the severity of a patient’s condition upon entry, accelerating critical care, such as Ramathibodi Hospital and Department of Mental Health. Hotels, such as HOP INN, are using the platform to develop AI concierges for hotel booking and financial services organizations like Ngern Tid Lor and the Government Savings Bank in Thailand are using AI sales assistants to nurture new customers. These AI assistants provide customer interactions that closely mimic those of local human experts at up to 95% accuracy.

“Cloud technology empowers organizations to run better and has enabled us to assist over 3,000 teachers nationwide in creating classroom modules that make education more accessible and effective,” said Dr. Winn Voravuthikunchai, founder of BOTNOI. “We leveraged AWS to build BOTNOI Voice, which delivers human-like interactions across multiple languages and diverse regional cultures.”

Dr. Winn Voravuthikunchai emphasized, “Amazon Bedrock allows us to seamlessly switch between foundation models for different use cases, which is critical for ensuring BOTNOI Voice meets the evolving needs of our rapidly growing client base and their discerning customers. We are thrilled to participate in the AWS Activate startup program, which has provided invaluable support to accelerate our growth on AWS. The program has enabled us to build, scale, and commercialize BOTNOI Voice more effectively.”

AI Tech Insights: SiMa.ai Launches Palette Edgematic on AWS Marketplace

  • BOTNOI’s culturally aware LLM uses AWS EC2 G4DN instances, the most cost-effective GPU instances for machine learning (ML) inference and graphics-intensive applications, for generating natural text to speech and speech to text synthesis audio. These instances provided scalable compute capacity in the cloud to handle the company’s growing generative AI workloads. BOTNOI plans to experiment with custom silicon Trainium2 in 2025 for up to 4x faster training than first generation Trainium chips.
  • Amazon Bedrock, the easiest way to build and scale generative AI applications with foundation models (FMs), powers BOTNOI’s generative AI assistant. This service gives the company the ability to easily access a choice of high-performing FMs from leading AI companies like AI21 Labs, Anthropic, Cohere, Meta, Mistral AI, Stability AI, and Amazon through a single API.
  • To filter out harmful content to the desired level, BOTNOI Voice uses Amazon Bedrock Guardrails to implement application-specific safeguards based on BOTNOI’s responsible AI policies, allowing them to configure content filter thresholds across categories including hate speech, insults, sexualized language, and violence. The startup also has plans to develop models on AWS that prevent the use of inappropriate or indecent words and unauthorized use or abuse of people’s voices on the platform.
  • Amazon SageMaker JumpStart, a ML hub with foundation models, built-in algorithms, and prebuilt ML solutions, helps BOTNOI train, finetune, and inference its model using local data speech, language, and cultural data gathered from consenting users.
  • BOTNOI integrates with Amazon Connect, an AI-powered contact center service that enables businesses to provide seamless customer service across various communication channels, to help organizations automate customer calls, screen incoming calls, and prioritize urgent interactions for rapid human intervention.
  • BOTNOI Voice is now available to millions of AWS customers on AWS Marketplace, a curated digital catalog designed to facilitate the discovery, purchase, deployment, and management of third-party software, data, and services that run on AWS. With the AWS Partner Network (APN), AWS Marketplace helps ISVs and Consulting Partners build, market, and sell their AWS offerings by providing valuable business, technical and marketing support.

BOTNOI is working with AWS Advanced Technology Partner, DoiT, to restructure the way their applications and systems are architected and deployed on AWS. With DoiT’s engineering team, BOTNOI has achieved over 75% savings on serverless computing and machine learning costs by making their cloud environment more efficient.

“The cultural diversity and distinct language needs across Southeast Asia present a unique opportunity for companies like BOTNOI to develop innovative, localized AI solutions,” said Vatsun Thirapatarapong, country manager, AWS Thailand. “We’re proud to support BOTNOI’s mission to make conversational AI more accessible. Their success demonstrates how AWS is empowering a diverse range of startups in the region to experiment big and achieve remarkable business growth.”

AI Tech Insights: AWS Announces New Database Capabilities Including Amazon Aurora DSQL, the Fastest Distributed SQL Database

Source – Amazon

To share your insights, please write to us at news@intentamplify.com