MUMBAI, India, June 26 -- Intellectual Property India has published a patent application (202621050432 A) filed by Moresh Madhukar Mukhedkar; Manish Jha; Shreya Jadhav; Soham Patil; Sandeep Kushwaha; Vivek Patil; and Shweta Koparde on April 20, 2026, for Modelmesh : A Decentralized Ai Marketplace.
Inventors include Moresh Madhukar Mukhedkar; Manish Jha; Shreya Jadhav; Soham Patil; Sandeep Kushwaha; Vivek Patil; and Shweta Koparde.
The application for the patent was published on June 19, 2026, under issue no. 25/2026.
Abstract: The present invention relates to a decentralized system for training artificial intelligence (AI) models using distributed computational resources and privacy-preserving techniques. The system integrates federated learning, blockchain technology, and decentralized storage to provide a secure, transparent, and efficient framework for collaborative AI model training. In the proposed system, AI developers submit models for training, which are distributed across multiple computing nodes contributed by participants. Training is performed locally on each node using federated learning, ensuring that raw data remains on local devices and only model updates are shared. This approach enhances data privacy and reduces the risk of data leakage. The system further utilizes blockchain-based smart contracts to record contributions and automate reward distribution in the form of tokens. This creates a trustless and incentive-driven environment where participants are compensated fairly based on their contributions. Additionally, decentralized storage using IPFS is employed to store trained models and related data securely. The invention addresses key challenges of centralized AI training systems, including high computational costs, data privacy concerns, lack of transparency, and inefficient resource utilization. By leveraging distributed computing and blockchain technology, the proposed system provides a scalable and cost-effective solution for AI model training.
Disclaimer: Curated by HT Syndication.
