
Tilde's AI assistants automate workflows and provide reliable support to customers and employees. They can be integrated with the organization's core systems (such as customer management solutions), providing access to personalized information and allowing users to perform specific actions during the conversation.
The MI assistant significantly reduces customer support costs, as evidenced by Rīgas namu pārvaldnieks, where savings reach up to 75% per customer conversation. In the case of Getliņi Eko, up to 80% of customer queries are resolved without human involvement, freeing up employee time. The Lithuanian Employment Service provides 24/7 customer service, where 40% of requests are resolved outside of working hours.
The MI assistant acts as a full-fledged member of the customer support team, providing personalized answers about products and services, available 24/7, and providing multilingual communication in the customer's chosen channel.
Not only does it answer questions, but it also uses the organization's knowledge and system integrations to guide the customer through processes.
Unlike standard chatbots, it optimizes resources, provides consistent communication, and allows access to all necessary information in one place, while maintaining control and ensuring scalability.
Configured for company research and risk evaluation work.
BTA Baltic Insurance Company, in collaboration with Tilde, developed Laura — a multilingual customer AI assistant designed to provide support in the field of health insurance.
"BTA Baltic Insurance Company, in collaboration with Tilde, developed Laura — a multilingual customer AI assistant designed to provide support in the field of health insurance. Laura is integrated into the BTA Latvia mobile app, where it provides instant, personalized answers about policies and their coverage, helping customers access the information they need quickly and securely. Health insurance-related questions are typically very specific, personal, and urgent. Customers need clear answers about what their policy covers, where to go, and how to proceed. Often, it is really important to make these decisions as quickly as possible. This means that until now, BTA employees had to handle a large volume of repetitive requests while providing accurate and personalized support. Operating in a highly regulated environment, every answer must be trustworthy, consistent, up-to-date, and aligned with company policies. Customizing customer support in such an environment is a real challenge. The solution was designed to fit into BTA’s existing ecosystem while meeting strict data accuracy and security requirements. The customer AI assistant is integrated into BTA’s policy management system, combining structured knowledge about the insurance industry with real-time customer data. All conversations are returned to BTA’s core systems, where they are stored securely. The assistant is available in the BTA Latvia mobile app. To use it, authentication is required, meaning only authorized users can access their information. The assistant has a scalable architecture to be ready for an increase in data volume, Laura combines the capabilities of large language models (LLMs), understanding of BTA’s health insurance, and instant access to data to provide quick and up-to-date answers. The assistant can: Explain policy coverage, limitations, and insurance conditions Show the remaining amount of insurance Check whether specific services and specialists are included in the coverage Answer general insurance questions using a structured knowledge base Connect the user with a support agent if necessary Customers can communicate with Laura both in writing and orally, making the service accessible to users with different needs. The assistant can be spoken to in three languages: Latvian, Russian, and English. Since the introduction of Laura, customer communication with BTA’s support service has changed. Laura can independently resolve around half of all incoming requests. The volume of calls related to health care questions has also decreased by around 40% at the Latvian branch, reducing the workload of customer support agents so they can focus on more complex issues. Customers now have continuous access to personalized information to conveniently get answers to questions related to insurance coverage and make the right decisions. The assistant also offers a simple feedback option so users can rate its answer with a thumbs up or down and add comments. This feedback is used to improve the quality of the assistant’s answers and the system’s performance based on users’ real needs. Currently, the customer AI assistant is available in the BTA Latvia mobile app, but the support provided to customers will be expanded in the near future by introducing integration into the Lithuanian app as well. This solution can be adapted over time to provide support in increasingly new situations, use additional sources of information, and evolve with changing customer needs and requirements. In the future, it is planned to expand support by adding additional insurance products as well as providing more policy-related information to customers. Laura undoubtedly proves that AI can improve customer experience in highly regulated industries by combining automation with reliable data. As demand for instant, personalized support continues to grow, this solution demonstrates what opportunities lie ahead for insurers and other organizations operating in a particularly complex, highly regulated environment."