STARTUP-STORIES
AI firm Fractal Launches Vaidya.ai, India’s First Advanced Medical LLM
SUMMARY
1. Fractal has launched Vaidya.ai, India’s first multimodal medical language model, designed to assist with diagnosis, treatment, and medical advice. The AI, trained on a vast dataset of images and text in multiple Indian languages, aims to provide reliable healthcare information while complementing, not replacing, medical professionals.
2. Vaidya.ai’s accuracy is validated using PG NEET exam questions, with an impressive 83% accuracy rate, placing it among the top 500. The AI integrates with Fractal’s existing healthcare product, Qure.ai, potentially offering a comprehensive enterprise solution for medical diagnostics and assistance.
Fractal, a leading AI firm, has launched Vaidya.ai, India's first multimodal medical language model (LLM) designed to assist users with diagnosis, treatment, and medical advice. This groundbreaking AI solution represents a significant leap in healthcare technology, providing a sophisticated tool to improve access to reliable medical information.
Vaidya.ai: Revolutionizing Healthcare Access
Vaidya.ai is built upon open-source base models, ranging from 30 to 70 billion parameters, and trained on a comprehensive dataset that includes over 650,000 images and 200,000 text inputs. This extensive training enables the model to understand multiple languages, including Hindi, Kannada, Tamil, and Telugu, making it a versatile tool for diverse user needs.
Srikanth Velamakanni, co-founder and CEO of Fractal, emphasized that Vaidya.ai is designed to complement rather than replace medical professionals. “AI is far from replacing doctors or making independent medical decisions. Our goal is to provide consumers with reliable healthcare information, offering a tool that enhances access and understanding,” Velamakanni explained.
Advanced Technology and Accuracy
Vaidya.ai stands out for its robust evaluation process. The AI’s accuracy is benchmarked using PG NEET exam questions, a highly competitive medical examination with 200,000 postgraduate candidates annually. Fractal’s AI has achieved an impressive accuracy rate, ranking among the top 500 with an 83% accuracy level, showcasing its reliability in providing medical insights.
Suraj Amonkar, Chief AI Research and Platforms Officer at Fractal, detailed the development process, highlighting the use of Large Language Models (LLMs) and Vision-Language Models (VLMs). “Our models are validated against 20+ years of PG NEET and PG Medical examination papers, ensuring their performance. While Vaidya.ai offers valuable insights, we always recommend consulting healthcare professionals for definitive medical advice,” Amonkar stated.
Integration with Existing Solutions
Alongside Vaidya.ai, Fractal continues to advance its existing product, Qure.ai, which specializes in detecting conditions like tuberculosis, lung cancer, heart failure, and stroke. The integration of Vaidya.ai with Qure.ai could pave the way for a unique enterprise solution in the healthcare sector, combining comprehensive diagnostic capabilities with advanced medical assistance.
Fractal’s commitment to AI in healthcare underscores its mission to enhance medical accessibility and accuracy through innovative technology. As Vaidya.ai progresses from a research project to a potential enterprise solution, it promises to transform how users interact with healthcare information.