Hi everyone,
We’re building Zant (Zig-Ant), an open-source SDK to make deploying Neural Networks on microprocessors easier. It’s written in Zig to maximize cross-compatibility and efficiency. The first version will take a NN model as input and outputs an optimized executable tailored for specific hardware. Long term, we aim to create a unified framework for model optimization and deployment.
Why another library?
• Many MCUs (e.g., ATMEGA, TI Sitara) lack robust deep learning libraries.
• No existing open-source solution lets you import, optimize, and deploy a model seamlessly.
• We’re using state-of-the-art optimization techniques inspired by Han Lab at MIT, we are starting academic collaboration with laboratories from Politecnico Di Milano (Italy’s top tech uni) and other institutions.
• Flexible enough to support new hardware with the same codebase.
If this excites you, give us a 
on GitHub and join us!
We’re looking for contributors—apply here: Contribute to Zant
__