TensorFlow inferencing using WebAssembly and WASI by @matei_radu radu-matei.com
As edge devices become more powerful, being able to perform inferencing on trained neural network models without recompiling application components for each architecture becomes important