MUMBAI, India, June 16 -- Intellectual Property India has published a patent application (202611054387 A) filed by Mr. Shubham Singh Mahar; Mr. Shivam Tomar; Mr. Shumoil Ansari; and Ms. Pranshi Verma, Ghaziabad, Uttar Pradesh, on April 29, for 'ai-driven nutrition & fitness system.'
Inventor(s) include Mr. Shubham Singh Mahar; Mr. Shivam Tomar; Mr. Shumoil Ansari; and Ms. Pranshi Verma.
The application for the patent was published on June 5, under issue no. 23/2026.
According to the abstract released by the Intellectual Property India: "This invention presents an intelligent, secure, and scalable AI-driven nutrition and fitness tracking system that uses Generative AI through the Google Gemini API combined with semantic similarity-based calorie estimation using transformer-based sentence embeddings. The system processes user-provided meal descriptions in natural language and delivers personalized, real-time dietary recommendations through a high-performance Golang backend and a React-based frontend interface. The proposed framework incorporates a sentence transformer model (all-MiniLM-L6-v2) for semantic food matching against the USDA FoodData Central database, a weighted cosine similarity retrieval mechanism for accurate calorie and macronutrient estimation, and a structured prompt engineering pipeline for AI-generated personalized nutrition insights via the Gemini API. The system integrates MongoDB for persistent user data storage, WebSocket-based real-time communication, JWT-based session management, OTP authentication, and a token bucket rate limiting algorithm for traffic stability anddenial-of-service protection. The system achieves 99.2% output consistency and a semantic similarity score of 0.94against trusted nutritional references including USDA FoodData Central and the Nutrition5k dataset. It maintains a median backend latency of 124ms, peak throughput of 820 requests per second, and WebSocket round-trip delay of 48ms. The architecture is designed for scalable real-world deployment while maintaining compliance with digital health security standards."
Disclaimer: Curated by HT Syndication.
