The operational process of NetsPresso®:

Trainer: The Trainer efficiently trains AI models using various datasets.

Compressor: The Compressor fine-tunes AI models to ensure smooth on-device functioning.

Converter: The Converter simplifies converting AI models into device-compatible frameworks.

Key Features

Efficient Model Tuning: Nespresso offers device-specific models to ensure optimal performance. Its open-source trainer allows for easy customization and is freely accessible to users. ​

Automatic Model Compression: The platform provides automatic compression techniques, including structured pruning and filter decomposition, to reduce model size and computational requirements without compromising performance. This facilitates deployment on devices with limited resources.

Framework Conversion and Quantization: NetsPresso supports the conversion of AI models into formats optimized for specific hardware. It enhances speed and efficiency by converting models to be device-friendly while maintaining their original performance. 

Accurate Model Benchmarking: The platform offers a cloud-based benchmarking service that executes models on real devices to verify accuracy and obtain performance metrics, including latency and memory usage. It supports over 20 different devices and provides results using various runtimes.

User-Friendly Interface: NetsPresso features a GUI client that enhances usability by providing functionalities such as code generation, project management, result visualization, and direct download of optimized models. ​

Financial Information:
In June 2024, Nota AI successfully raised $19.9 million (25.8 billion KRW) in its latest Series C funding round. This brings the total investment raised since its founding to approximately $42.6 million (53.2 billion KRW). The round was led by STIC Ventures and LB Investment, with new investments from STIC Ventures, Korea Development Bank, and Mirae Asset Securities.