From Data to Deployed Edge AI
From data collection to on-device deployment, AI Tool Stack keeps the workflow in one controllable system.
Built for NE301. Open for all.
Not a Training Interface.
A Deployment Pipeline.
AI Tool Stack is not another training interface. It is a system that takes models from raw data to real edge devices.
One workflow. Fully under your control.
Run the entire pipeline on your own infrastructure. No cloud dependency, no third-party data transfer.
Fully open on GitHub. Modify, extend, and integrate — without vendor lock-in.
Optimized for NE301 from quantization to deployment. No adaptation layer required.
Docker-based setup, standard formats, and a workflow you can inspect and control end-to-end.
From Data Ingestion to Device Deployment
Self-Hosted. Browser-Based.
No local installs beyond Docker. No local setup beyond Docker. The pipeline runs entirely in a self-hosted web interface — covering data, training, and deployment.
NE301 Closes the Deployment Loop
Quantization and deployment are built around the NE301's STM32N6 Neural-ART NPU. If you're targeting edge AI that runs entirely on-device — NE301 is where the pipeline closes.
Build Your Edge AI Pipeline.
Start today.
Open source, self-hosted, no lock-in.
The full experience is one NE301 away.
Questions? Join us on Discord →