We developed 2B Well as the platform for psychological support to refugees affected by military conflicts. Our main aim is to help those who have experienced the traumatic impacts of the war. We provide them the access to professional psychological care and resources such as one-on-one counseling, group therapy sessions and mental health AI chatbot. The future vision for expanding the product is to provide mental health support by leveraging AI capabilities to support people from other countries that may be affected by military conflicts in the coming decades.
Client needs
A significant challenge in the context of mental health support within regions affected by military conflict is the language barrier; most of the available mental health support applications are primarily in English. This situation is compounded by the fact that standard psychological services in these regions have become overwhelmed due to increased demand. Consequently, there is an urgent need for innovative solutions that can resolve mental health problems on a large scale. These solutions must be both understandable and accessible to the majority of the population to offer support in their native languages and adapt to the unique cultural and societal contexts of the affected regions.
Discovering business challenges
The project faced several technical challenges such as adapting AI algorithms to effectively understand and respond to the cultural nuances of Ukrainian users. It required extensive training on relevant datasets. Additionally, designing the infrastructure to scale horizontally with fluctuating user demand while maintaining performance was crucial. As well the app was maintaining compliance with data protection regulations for sensitive mental health information, which involved stringent security measures. The integration of data analytics for real-time user experience posed challenges in data collection. In addition, the maintenance of AI models required a complicated setup to support smooth operations.
Development of AI scientific based solutions
The discovery phase of 2BeWell began with a deep dive into understanding the challenges faced by refugees. They often face systemic barriers such as language difficulties, limited access to qualified therapists, and mental health solutions without cultural backgrounds. Many expressed feelings of isolation and a lack of accessible resources to resolve their mental health needs.
Through interviews and research, refugees voiced a strong desire for immediate, empathetic, and culturally appropriate mental health support. Professionals in the field highlighted the importance of adapting care to reflect cultural nuances. So, it’s get clear that the solution required more than just traditional therapeutic support. There was a strong need for a scalable, personalized system that could connect users to immediate care while alleviating the pressure on therapists.
Integration of web and mobile applicatins
We integrated a dual-platform solution to combine advanced AI and human expertise. The project introduced Uncle Zen, an AI-powered chatbot that provides real-time psychological assistance in multiple languages with diverse cultural contexts. This chatbot was complemented by a user-friendly mobile application, which allowed individuals to access support instantly, connect with volunteers and engage with professional therapists.
On the other side, a complex web portal was developed for psychotherapists and volunteers. This platform facilitated effective case management, secure communication and the delivery of personalized care. The integration of these tools created a seamless interaction between users and professionals to maintain a holistic mental health support system.
Prototype designing
The design of the prototype focused on usability, inclusivity and scalability. The mobile and web applications were developed with strong intuitive navigation to help users with different levels of technological familiarity. Personalization was a key feature with AI adapting interactions based on user preferences and needs. The system was also built to handle a large user base while providing high-quality and responsive care.
The prototype blends cutting-edge technology with the human touch to resolve the unique needs of refugees and empower professionals to deliver culturally sensitive and effective mental health care.
ZenBit team is ready to assess, design, build, modernize, and optimize your healthcare IT solutions and infrastructure to help you run daily IT operations.
Our mental support solution for resilience
The aim of 2BWell platform create a mental health support system tailored to meet the unique needs of refugees. The solution integrates AI, advanced software development and data-driven experience to deliver culturally adaptive care.
At the heart of the platform is Uncle Zen, an AI mental health chatbot developed with Hugging Face’s Transformers based on Anthropic Claude LLM. The chatbot was fine-tuned to understand the cultural nuances and mentality specific to Ukraine. It provides users with empathetic, instant mental health support in their native context.
To facilitate real-time communication, we developed a mobile application with React Native to ensure compatibility across iOS and Android devices. The app connects users with trained volunteers and psychotherapists through a reliable chat and video or audio conference powered by WebSockets. An admin panel for volunteers, built with ReactJS and NodeJS (NestJS). A scalable infrastructure design incorporates BULL MQ for efficient queue management and Redis for WebSocket session storage to enable horizontal scaling to handle any workload with the ability to automatically descale when demand decreases.
We also implemented DevOps, MLOps and FMOps principles to automate the deployment of both infrastructure and AI models. This includes automated training, fine-tuning, validation and quality checks to reduce deployment times by up to 80% with deployments now taking approximately 15 minutes.
To continually improve the user experience, we integrated the Event and Property data analytics concept based on the solution described in the article “Best practices to implement analytics for your digital product”. This solution enables us to collect user behavior data, allowing us to make informed decisions about new features and improvements. By relying on real-use data, we can ensure that future updates are more relevant and aligned with user needs.
The 2BeWell platform demonstrates a holistic approach to mental health care. By combining AI innovation, real-time human support and data-driven development, it bridges the gap between individuals in crisis and the qualified mental health care they need. This scalable, culturally adapted solution sets a new benchmark to resolve mental health challenges.
Conclusion
The 2B Well platform delivers personalized and culturally relevant support to refugees. Through innovations like the Uncle Zen chatbot, cross-platform mobile applications and scalable infrastructure, 2B Well not only ensures accessibility but also fosters a sense of safety and connection for users. The platform’s advanced analytics and automation processes empower continuous improvement for adapting to audience needs.
The success of this project is owed to the key contributions of Prof. Richard Barker, Tania Orlova and Dmitry Broshkov, whose vision and inspiration has been fundamental to 2B Well’s achievements.
healthcare, web platform, telemedicine, toolkit app
