Featured Stories

Congratulations to Nine Mr. Natthanarong Tiangjit, a Tech Talent scholarship student from the School of Engineering, who joined the team as part of winning the 1st place championship award for the Eastern Coordination Center in the Super AI Engineer Season 5 program, AI Innovator Track, with the innovation "RaoChatHub" - an intelligent chat platform that uses real databases, capable of reducing costs by up to 20 times and has the potential to develop into a startup in the future.
 

Starting Point from Competition

The beginning of this project came about after team members presented their work and passed the regional selection, at which time the work was still in the prototype stage. Having previously worked together in competitions the year before, they invited Nine to join the team to further develop the system to make it fully functional. This competition divides the areas into 5 coordination centers: North, Northeast, West, East, and South. The first-place winners from each region are all considered national champions and receive the right to present their work at the Super AI Engineer: The 5th National AI Exhibition.

Strong 6-Person Team from Multiple Institutions

The team consists of 6 members with clearly divided responsibilities, including Marketing, Business, Developer, and AI divisions, with members from several universities including Burapha University (Project Manager) and Silpakorn University (Design and User Journey). "The work took 2 months with regular meetings and progress tracking, just like real work."

What is RaoChatHub?

RaoChatHub is a chatbot that differs from ordinary ones. Instead of having AI answer questions generically, it uses a Knowledge Database system to retrieve information from actual documents.

Key features of the system:

1. Uses Real Data, No Guessing - Users can upload PDFs, documents, or various data into the system. The chatbot then searches for answers from these databases. If no answer is found, it forwards the question to an admin to answer instead.
2. Can Read Images with Vision AI - Can read screenshots that users capture and send, which ordinary chatbots cannot do.
3. Analyzes Real Intent - The system can understand context and meaning of questions, even if users ask multiple questions together or use unclear words.
4. Consolidate data connection channels - gather messages from Facebook, Instagram, and Line into a single system. Admins don't need to respond across multiple channels, and most importantly, reduce costs by up to 20 times. Ideal for SMEs and medium-sized businesses."
5. Reduces Costs by 20 Times - Suitable for SMEs and medium-sized businesses.

Solving Critical Problems of Ordinary Chatbots

This project originated from real problems with current chatbots that often provide generic answers and sometimes give incorrect information. For example, there was a case where an airline's chatbot provided wrong information about ticket discount policies, resulting in the company being held liable according to a court ruling. A bank in America once laid off employees and replaced them with a general chatbot but had to rehire staff within a few months due to ineffective responses. RaoChatHub solves this problem by using only real data and has a system to transfer to human agents in cases where questions are complex or users want to talk to a real person. RaoChatHub is ready to use now. Currently, the system has been implemented and further developed at the Computer Center, Burapha University, to answer students' questions 24 hours a day. It helps reduce the burden of answering repetitive questions from staff, such as course registration, password recovery, and Wi-Fi connection, allowing staff to have more time to handle more complex issues.

RaoChatHub is Ready for Use Now

         Currently, the system has been implemented at the Computer Center, Burapha University, to answer students' questions 24 hours a day, helping reduce the burden of answering repetitive questions from staff, such as course registration, forgotten passwords, and Wi-Fi connection, allowing staff to have more time to handle more complex problems.
 

Challenges and Lessons

Nine shared the challenges and new lessons discovered: "I was mainly responsible for the AI aspect. The biggest challenge was communicating with multiple departments because each department has different needs and understanding. Marketing might not understand why development takes time, while the technical team needs to make the system actually work. This project isn't just about writing programs, but requires understanding both Business and Marketing as well."

Next Step Towards Startup

The team's future after receiving initial funding of 10,000 baht from the Super AI Engineer project and preparing to apply for additional funding from TED Fund to further develop into a serious startup: "The future plan is to develop the system to be able to close sales as well, not just answer questions, suitable for SMEs that need to answer questions and sell products simultaneously. The goal is for the chatbot to communicate naturally like a real person."