AxonML -- A PyTorch-equivalent ML framework written in Rust

Andrew Jewell Sr / AutomataNexus LLC Over the past year and a half, I've been building AxonML -- a machine learning framework in Rust that aims for feature parity with PyTorch. It's now at v0.3.2: 22 crates, 336 Rust source files, 1,095 passing tests, and it's running production inference on Raspberry Pi edge hardware in commercial buildings. This post covers why I built it, how it's architected, the hard technical problems I ran into, and where it's actually being used. GitHub: github.com/AutomataNexus/AxonML License: MIT / Apache-2.0 Motivation I built an entire building automation ecosystem from scratch. NexusBMS is the central building management platform -- won an InfluxDB hackathon with it, runs InfluxDB 3.0 OSS alongside my own database (Aegis-DB, also open source). The edge controllers are 50+ Raspberry Pi 4/5s running my custom NexusEdge software: Rust hardware daemons for I2C, BACnet, and Modbus communications, direct HVAC equipment control via analog outputs, 24V triacs, 0-1