- Jivi MedX, developed by Indian startup Jivi, secures top rank on Open Medical LLM Leaderboard.
- Surpasses established LLMs like OpenAI’s GPT-4 and Google’s Med-PaLM 2 with an average score of 91.65.
- Evaluation conducted by Hugging Face, University of Edinburgh, and Open Life Science AI, covering medical exams and research questions.
- Jivi aims to revolutionize primary healthcare through generative AI, offering high-quality care at reduced costs.
- Utilizes extensive proprietary medical dataset and instruction fine-tuning algorithm for training Jivi MedX LLM.
Main AI News:
Jivi, an Indian healthcare AI startup founded by Ankur Jain, former BharatPe Chief Product Officer, and GV Sanjay Reddy, Chairman of Reddy Ventures, has achieved the remarkable feat of securing the number one rank on the Open Medical LLM Leaderboard with its purpose-built medical Large Language Model (LLM), Jivi MedX. Surpassing industry giants like OpenAI’s GPT-4 and Google’s Med-PaLM 2, Jivi MedX attained an impressive average score of 91.65 across all nine benchmark categories of the leaderboard.
This accolade comes as a testament to Jivi’s dedication to revolutionizing primary healthcare through the application of generative AI, providing high-quality care round-the-clock at a fraction of the conventional cost. Ankur Jain, CEO of Jivi, expressed the company’s mission to utilize artificial intelligence to elevate patient care standards, emphasizing how their platform expedites diagnostics and ensures heightened accuracy, thereby facilitating timely and precise treatment for patients globally.
Facilitated by leading AI platform Hugging Face, in collaboration with the University of Edinburgh and Open Life Science AI, the Open Medical LLM Leaderboard serves as a comprehensive assessment platform for medical-specific LLMs, evaluating their efficacy in answering a wide array of medical questions sourced from exams and research. The evaluation spans various medical examinations such as the AIIMS and NEET in India, as well as the USMLE, alongside comprehensive assessments covering clinical knowledge, medical genetics, and professional medicine, among other domains.
Jivi, currently operating with a lean team comprising 20 members, including physicians, surgeons, AI engineers, and data scientists, is committed to leveraging technology to revolutionize healthcare accessibility, affordability, and quality on a global scale. Central to Jivi’s success is its vast proprietary medical dataset, comprising millions of medical research papers, journals, clinical notes, and other authoritative sources, which serves as the foundation for training the Jivi MedX LLM. Notably, this dataset stands among the largest of its kind globally. Furthermore, Jivi MedX underwent training leveraging an instruction fine-tuning algorithm known as Odds Ratio Preference Optimization (ORPO), underscoring the company’s commitment to leveraging cutting-edge methodologies in its pursuit of excellence in healthcare innovation.
Conclusion:
Jivi’s success in claiming the top spot on the Open Medical LLM Leaderboard signifies a significant advancement in healthcare AI technology. This achievement underscores the potential for AI-driven solutions to enhance patient care standards and revolutionize the accessibility, affordability, and quality of healthcare globally. As Jivi continues to innovate and expand its reach, it poses a formidable challenge to established players in the market, setting a new benchmark for excellence in the field of medical language modeling.
