Customer Overview

PT Beranda Wawasan Konsumen (AcuityHub) is a technology company that provides an online platform for conducting surveys and market research. Their platform allows businesses and organizations to create and distribute surveys to a panel of respondents, enabling them to gather valuable insights and data for their research needs. AcuityHub plays a crucial role in helping businesses and organizations make informed decisions by gathering and analyzing data about consumer behavior, preferences, and trends. The market research industry has been rapidly evolving, with an increasing demand for real-time insights and the adoption of advanced technologies to streamline data collection and analysis processes.

Business Challenge

The AcuityHub Muniyo users often struggled to come up with effective survey questions and were unaware of the platform’s full capabilities. While the Muniyo app had an FAQ section to help users understand the platform, it was not effective in solving user problems. As a result, users were unable to create the best possible surveys, leading to suboptimal data and insights.

Technology Challenge

AcuityHub needed a solution that could provide Muniyo users with personalized assistance in creating surveys and understanding the platform’s features. Traditional rule-based chatbots and FAQs were not sufficient to address the diverse and complex needs of users. AcuityHub required a more advanced and intelligent system that could understand natural language queries and provide relevant and contextual responses.

Solution: Generative AI-Powered FAQ Chatbot

ICS Compute’s solution is to create a virtual assistant powered by AWS Bedrock, so the Muniyo users get some insights and information from GenAI Virtual Assistant based on the knowledge base or via chat interface.

The FAQ Chatbot Powered with GenAI solution has benefited over 1,000 active users of the Muniyo platform, enabling them to quickly access information and engage with the platform more effectively. By reducing the need for manual customer support and increasing customer engagement, AcuityHub has realized cost savings annually, while also improving customer satisfaction and retention rates.

The solution was built using Amazon Bedrock, leveraging its advanced natural language processing capabilities and few-shot prompting techniques. The solution was using AcuityHub’s domain-specific data, enabling it to provide accurate and relevant responses to customer inquiries.

The solution leveraged a few-shot prompting by providing the Claude v3 Haiku and Sonnet models with a small number of examples from AcuityHub FAQ documents. This allowed the model to quickly understand the context and generate relevant responses. Additionally, an advanced Retrieval Augmented Generation (RAG) implementation was employed, which enabled the model to retrieve and incorporate relevant information from AcuityHub’s knowledge base during the summarization process.

Cohere’s Multilingual Embedding Model v3 was used to generate vector representations of AcuityHub knowledge base, which included both Bahasa Indonesia and English content. These vector representations were then stored in the Amazon Aurora PostgreSQL database with the pgvector plugin, enabling efficient retrieval of relevant information during the summarization process.

On Responsible AI

Following ICS Compute’s Responsible AI Policy, we have designed this use case for AculityHub’s platform, prioritizing the principles of Transparency and Reliability & Safety.

In line with the Transparency principle, our use case ensures clear user involvement and oversight throughout the process. Users maintain control by logging in, submitting queries, and checking the Quality Check (RAG) result. We provide clear documentation and communication about the AI’s role and limitations, fostering transparency in its involvement. This approach aligns with our commitment to pursuing interpretable AI solutions with human oversight.

Additionally, we have implemented a Quality Check step to uphold the Reliability & Safety principle. This step ensures rigorous testing and validation, guaranteeing that the AI operates safely and as intended, providing accurate and reliable responses to user queries. By prioritizing these principles, ICS Compute aims to deliver trustworthy and transparent AI solutions that augment user experiences while maintaining human control and oversight.

Customer Outcomes And Benefits

1. Increased User Engagement and Satisfaction: By providing personalized assistance and addressing user queries effectively, the virtual assistant has improved user engagement and satisfaction. According to surveys conducted by AcuityHub, 63.3% of repeat users (38 of 60 surveyed users) have reported an improvement in their user experience .
2. Cost Savings: The virtual assistant has reduced the need for manual customer support, resulting in cost savings for AcuityHub. Users can now access information and assistance 24/7 without the need for human intervention.
3. Improved Survey Quality: By guiding users in creating effective survey questions and leveraging the platform’s full capabilities, the virtual assistant has contributed to the creation of higher-quality surveys, leading to more valuable insights and data for AcuityHub’s customers.
4. Scalability: The Generative AI-powered virtual assistant can handle a large volume of user queries simultaneously, ensuring that AcuityHub can scale its operations without compromising on customer support and assistance.

According to Setyo, AcuityHub Deployment Team Leader, “Our customer support operations have been completely transformed by the GenAI Chatbot. The customer support virtual assistant has reduced the workload for our human agents by handling a variety of inquiries quickly and accurately, freeing them up to work on more difficult surveys and market research. This has improved our overall operational efficiency and productivity.”