As of July 1, 2024, Bonsai d.o.o. has been formally merged into Span d.d. Upon the registration of the merger with the Commercial Court in Zagreb, the merged company ceases to exist, and Span d.d. becomes its universal legal successor. The integration of Bonsai into Span will unify Span's software development offerings.
The bonsai.tech website will be permanently shut down. We thank you for your interest and support over the years. For all additional information and updates, please visit www.span.eu and their social media channels.
EOS Matrix was looking for a solution that would streamline communication with their clients by automatically answering their frequently asked questions. They wanted to provide instant support to their clients without the need for the presence of debt/billing experts.
Aware of the fact that information distributed through a chatbot is confidential and private, a solution had to be able to verify users to provide them with personal debt information that cannot be shared with a third party.
EOS Group is one of the leading technology-driven financial investors and an expert in the processing of outstanding receivables. The company's core business is the purchase of unsecured and secured debt portfolios. With over 40 years of experience, EOS offers some 20,000 customers in 26 countries around the world smart services for all their receivables management needs.
Here is a more detailed overview of this case study. It might serve you as an inspiration for your own chatbot.
Leveraging a chatbot solution allowed EOS Matrix to rapidly address incoming messages. Using it to provide information and collect feedback, EOS Matrix recorded a decrease in the workload of customer support agents. This went a long way to reduce expenses while enriching the customer experience.
EOS Matrix web site was enriched with a virtual assistant called Lea whose tonality is tailor-made to match the company image and retain wanted customer perception. She can answer questions regarding EOS as a company and their work, as well as to provide information about individual debts. She also collects customer feedback on EOS and chatbot itself so those areas can get improved over time.
Lea comes with the web admin interface that helps with the chatbot’s supervision. Chatbot usage analytics provides valuable insight into conversations, and conversation history can help in case of a misunderstanding. Gathered unanswered questions keep admins on track with new frequent queries and results of chatbot questionnaires gives an overview of collected feedback.
Lea greets her visitors and leads simultaneous conversations focusing on the received replies. Users can also ask Lea questions, upon which she searches her knowledge base and provides the best-known answer.
This virtual assistant is able to identify the user and give detailed information about personal debts. Identity verification is carried out through two-factor authentication for higher security of confidential data. For the purpose of user verification, integration with EOS’s CRM had to be made.
Another integration scenario is the one when Lea can’t find the answer in her book of knowledge and offers the user to request a call from her real-life colleagues. In that case, task for an agent call is open in EOS’s CRM. Furthermore, few more integrations were made to enable Lea to display information such as detailed financial debt printout and payment details to users.
As mentioned, a part of chatbot solution is an admin application which main purpose is a chatbot supervision. The application has user management functionality which enable access control to certain parts of the application. For example, Agent user can access usage statistic, see conversation history, but it doesn’t have edit rights on the chatbot knowledge base. This functionality is available for the Admin users.
Chatbot’s knowledge base contains a list of QnA pairs that cover conversational topics. To improve chatbot’s ability to answer a certain question correctly, multiple question can be defined for each answer. Our custom-made language processing QnA service determines which answer will be served through the bot.
The overall chatbot solution is designed to match the EOS Matrix’s operations and industry standards. Through continual communication, bonsai.tech made sure to understand EOS’s needs, share its vision and deliver the desired solution.
Due to flexibility in the development process, integration with EOS internal systems and databases were also established to ensure the highest business value. As the EOS Project Manager said: “bonsai.tech showed professional approach, prompt reactions and good problem solving skills that all lead to effective finalization of EOS chatbot project.”
bonsai.tech came across as a reliable partner experienced in automating business processes and delivering business value. They designed a chatbot that successfully meets our needs.
- Denis Jakob, Business Analyst, EOS Matrix
