KDAN Accelerates Development of Generative AI Applications by 50% with Amazon Bedrock

Taiwan-based KDAN develops software-as-a-service (SaaS) solutions for digital document management and eSignature products and services. Launched in 2009, the company also operates in China, the US, Japan, South Korea, and Singapore. Its software products boast over 12 million registered members, including over 40,000 business customers. In 2023, the Financial Times recognized KDAN among the Top 500 High-Growth Companies in the Asia-Pacific region.

KDAN’s target customers are increasingly interested in artificial intelligence (AI)-powered applications. Ray Lee, special assistant to the chief operating officer at KDAN, comments, “The public sector in Taiwan, where we’re aiming to grow, is looking to work smarter and deliver better services to citizens through AI-driven solutions. Organizations around the world recognize the potential of AI technology to improve processes.”

To stay ahead of competitors, KDAN set out to accelerate its generative AI program. The company saw great opportunities for generative AI to improve its document management and eSignature applications, automating tasks like document analysis and data extraction and providing new insights. Additionally, it could develop chatbots to compare information and make document workflows more efficient.

To go to market faster, KDAN turned to Amazon Web Services (AWS), which has supported its SaaS products from the beginning. “We can rely on AWS to provide the scalability, security, and cost-effectiveness needed to maximize our innovation capabilities in the generative AI space,” states Lee.

KDAN joined a series of AWS workshops and training sessions to learn best practices for AI and generative AI workloads on AWS. Following this, work began in earnest. AWS AI experts helped KDAN integrate Amazon SageMaker into its AWS stack to build, train, and deploy its machine learning (ML) models. Additionally, they collaborated with KDAN’s team to implement Amazon Bedrock for running large language models like Llama 2 and Claude 3.

“We used Amazon SageMaker's robust machine learning capabilities and Amazon Bedrock's scalability to enhance our ability to develop and deploy advanced AI solutions efficiently,” explains Lee. He goes on to explain that his team was impressed by the collaborative spirit at AWS: “We liked their joint problem-solving approach. From the outset, AWS worked closely with us to develop customized solutions tailored to our specific use cases, ensuring our generative AI implementation met our business objectives.”

During the trial, KDAN developed a generative AI–powered intelligent document analyzer application for data analysis and extraction. It also created generative AI chatbot assistants for contract and customer relationship management. “The chatbot provides employees with personalized answers on our internal processes and a copilot solution to streamline the production of code,” Lee explains.

By using Amazon SageMaker and Amazon Bedrock and collaborating with AWS, KDAN halved the normal development time for its generative AI applications. “Working with AWS accelerated the development of our generative AI applications by at least 50 percent,” says Lee. “Launching into the market faster helps establish KDAN as a leading local provider of generative AI solutions in our field.”

With its generative AI applications running on AWS, KDAN can drive expansion across domestic and international markets. “Once we’ve solidified our success in Taiwan, we’ll showcase our technology in core markets like the US,” Lee says.

KDAN has launched a series of proofs-of-concept for its generative AI–powered PDF Reader with public-sector agencies in the local market. End users include environmental engineering consultants, graduate students, and human resources (HR) personnel.

“Engineering consultants have used our generative AI application on AWS to summarize key points in foreign documents, increasing reporting efficiency by 75 percent, and students have used the solution to search out the more relevant information for their project,” says Lee. “For HR staff, it’s helped speed up job application screening processes by 70 percent.”

The company is looking forward to continuing generative AI development using AWS. With technology advancing so quickly, KDAN believes customer benefits will increase exponentially. “We want to be at the forefront of developments in AI technology, differentiating our offerings in a competitive market by helping customers unlock even more value from growing amounts of data,” Lee concludes.