TinyML algorithms are introduced for managing Big Data in large-scale IoT systems. These algorithms—TinyCleanEDF, EdgeClusterML, CompressEdgeML, CacheEdgeML, and TinyHybridSenseQ—work together to enhance data processing, storage, and quality control in IoT networks using Edge AI. TinyCleanEDF applies federated learning for data cleaning and anomaly detection. EdgeClusterML combines reinforcement learning with self-organizing maps for effective data clustering. CompressEdgeML uses neural networks for adaptive data compression. CacheEdgeML employs predictive analytics for smart data caching, and TinyHybridSenseQ focuses on data quality evaluation and hybrid storage strategies. Experimental results show that these algorithms outperform similar methods across various evaluation metrics. The study highlights the importance of TinyML in efficiently managing IoT data, reducing reliance on central systems, and minimizing data transmission needs. TinyML enables localized data processing, enhancing network efficiency, security, and reliability. The proposed algorithms are evaluated using Raspberry Pi devices, demonstrating their effectiveness in real-time IoT scenarios. The study emphasizes the role of TinyML in addressing Big Data challenges in IoT, including storage, processing, and security. The framework includes a centralized system for data aggregation and analysis, with feedback mechanisms for continuous optimization. The hardware setup includes Raspberry Pi devices and various sensors for data collection. The computational framework involves a centralized HPC cluster and Raspberry Pi units for distributed processing. The dataset includes environmental, motion, light, distance, and soil moisture data, providing a comprehensive basis for algorithm testing. The proposed algorithms are designed to handle specific challenges in IoT systems, ensuring efficient and secure data management. The study concludes that TinyML offers a comprehensive solution for managing Big Data in IoT, enhancing efficiency, security, and scalability.TinyML algorithms are introduced for managing Big Data in large-scale IoT systems. These algorithms—TinyCleanEDF, EdgeClusterML, CompressEdgeML, CacheEdgeML, and TinyHybridSenseQ—work together to enhance data processing, storage, and quality control in IoT networks using Edge AI. TinyCleanEDF applies federated learning for data cleaning and anomaly detection. EdgeClusterML combines reinforcement learning with self-organizing maps for effective data clustering. CompressEdgeML uses neural networks for adaptive data compression. CacheEdgeML employs predictive analytics for smart data caching, and TinyHybridSenseQ focuses on data quality evaluation and hybrid storage strategies. Experimental results show that these algorithms outperform similar methods across various evaluation metrics. The study highlights the importance of TinyML in efficiently managing IoT data, reducing reliance on central systems, and minimizing data transmission needs. TinyML enables localized data processing, enhancing network efficiency, security, and reliability. The proposed algorithms are evaluated using Raspberry Pi devices, demonstrating their effectiveness in real-time IoT scenarios. The study emphasizes the role of TinyML in addressing Big Data challenges in IoT, including storage, processing, and security. The framework includes a centralized system for data aggregation and analysis, with feedback mechanisms for continuous optimization. The hardware setup includes Raspberry Pi devices and various sensors for data collection. The computational framework involves a centralized HPC cluster and Raspberry Pi units for distributed processing. The dataset includes environmental, motion, light, distance, and soil moisture data, providing a comprehensive basis for algorithm testing. The proposed algorithms are designed to handle specific challenges in IoT systems, ensuring efficient and secure data management. The study concludes that TinyML offers a comprehensive solution for managing Big Data in IoT, enhancing efficiency, security, and scalability.