Healthvana, a patient engagement platform company, is revolutionizing the patient experience with its unique approach. The company aims to reach one million patients by the end of 2024, and it is already serving tens of thousands of patients. Ramin Bastani, CEO of Healthvana, emphasized the importance of digitizing the patient experience and addressing the flaws in patient portals, which are generally disliked by patients. Healthvana’s platform, one of the largest patient-facing AIs in the U.S., offers instant responses connected to patients’ health histories and past chats.
One of the standout features of Healthvana’s platform is the option for patients to interact with a drag queen AI persona. This persona creates a fun, non-judgmental environment that encourages open conversations between patients and the AI. Surprisingly, 80% of patient conversations are happening with the drag queen persona, indicating that patients are more comfortable engaging with this persona. Additionally, 25% of messages occur outside of clinic hours, allowing Healthvana to engage traditionally hard-to-reach patients on their own terms.
Healthcare providers struggle to keep up with the influx of patient messages, making AI assistance crucial. Healthvana’s platform underwent rigorous testing and evaluation with the help of more than 100 clinicians. The company also worked closely with OpenAI to refine its AI models and ensure accurate, comprehensive, and contextually sensitive responses. Healthvana’s evaluation process involves human content moderators reviewing all messages within 24 hours and rating them based on accuracy and comprehensiveness. Anything that doesn’t meet the highest rating is put into a product feedback loop for improvement.
The evaluation process consists of three categories: model evaluation, single-turn evaluation, and multi-turn or multi-channel application. Model evaluation focuses on finding the best model for solving users’ problems. Single-turn evaluation ensures that the AI agent can effectively address users’ needs. Multi-turn evaluation involves back-and-forth conversations between the AI and the user to refine the conversation. Multiple AI models interact with each other on the back end, using various tools and databases to enhance the patient experience.
Healthvana’s commitment to thorough evaluation and constant improvement sets it apart from other platforms. The company’s dedication to delivering accurate and contextually sensitive responses is evident in its human review process and continuous refinement of the message evaluation model. By leveraging AI technology and incorporating a drag queen persona, Healthvana is making healthcare more accessible, engaging, and enjoyable for patients.