Lavita 2025 Development Roadmap
When Lavita AI launched just weeks before the start of 2023 (link), we set out with a bold vision: a world where people are empowered with full control of their health data, with no barriers to harnessing it, and where they can securely access the latest medical knowledge and AI models to generate personalized insights that improve their health and well-being.
Along this vision, Lavita is building a trusted platform where researchers conduct cutting-edge medical and AI research using a decentralized-first approach to collect data, access compute resources, and validate and share results with communities — ultimately pushing the boundaries of science in medicine.
Furthermore, LAVITA token serves as the foundational incentive layer to complete the economic flywheel by unleashing inherent the value of health data for its owners (i.e., Individual Ownership of Data) to fuel access to the most advanced AI models and enable a decentralized health data platform for medical research.
Today, we are closer than ever to making that vision a reality and are excited to share with you the 2025 Lavita development roadmap:
Secure AI Companion™
As previously announced in Lavita’s presentation (link) during the ThetaCon’24 conference, Lavita will soon launch its flagship mobile app on iOS and Android platforms. Using the app, users will be able to seamlessly connect their health data (including Electronic Health Records, test results, medical imaging, wearables data, etc.) and interact with their data using privacy-preserving AI foundation models.
Over time, even though AI foundation models — including general-purpose ones — have advanced in answering general medical and clinical questions, they still lack the ability to incorporate users’ personal medical records. More importantly, sharing sensitive health data with third-party apps and APIs raises serious privacy concerns, as users often have no control over how their sensitive information is used. And that’s exactly the problem we solve at Lavita.
Lavita’s implementation of confidential computing (link) brings some of the most advanced open-source models from Llama, DeepSeek and others to users’ fingertips, enabling them to access reliable, secure, and personalized medical information. Security and data privacy are at the heart of the Lavita platform, ensuring that user data remain secure at every step of engagement with the platform. We are happy to announce that Lavita is bringing one of the most advanced chips on the market, the NVIDIA H100 Tensor Core GPU, to its confidential virtual machine implementation.
In addition to using the mobile platform to access personalized medical insights, users will be able to receive LAVITA rewards for Staking Data on the platform, to participate in research studies. Participation in such research studies is completely voluntary and through informed consent for the users. When opted-in, users can securely share their health and medical data with paying researchers, while ensuring privacy and security of their data through each step of the research study. The platform revenue from such research studies will be shared back with the participating users through token buybacks, to translate Data Velocity into Token Velocity.
Medical Evaluation Sphere™
Conducting high-quality testing and evaluation is a critical part of building any robust AI product or pipeline. In the medical and clinical domains, this becomes even more important due to the sensitive nature of real-world tasks that directly impact users’ health and well-being.
Looking at the current landscape, we identified a major gap between the existing tools for rigorously testing AI models in healthcare and the urgent need for high-quality evaluations in real-world medical and clinical tasks. That’s why we launched the Medical Evaluation Sphere — to bridge this gap. We believe that achieving high-quality, realistic evaluations requires fostering a global community of medical experts and users — something Lavita is uniquely positioned to build. Comprehensive evaluations not only help us continuously monitor and assess the quality of rapidly evolving AI models in medical and clinical applications, but they also provide critical insights for building better models that address the challenges we care about.
We’ve already seen growing engagement from medical professionals and users eager to participate in the Medical Evaluation Sphere. In 2025, we are doubling down on our efforts to expand both this network and its engagement, establishing the Medical Evaluation Sphere as the go-to platform for real-time evaluation of foundation models — across any medical or clinical task, in any modality, and at scale.
Lastly, we’d like to thank the Lavita community and supporters for trusting our team to bring cutting-edge AI and Confidential Computing technologies to healthcare. Now more than ever, we’re excited to contribute to improving health and wellness worldwide.
