Nani'
research, ventures, code

Nani'

Research notes, experimental systems, and product prototypes. That is the whole split.

01
Research Speech, generative audio, technical reports, and longer-form experimental work.
↗
02
Ventures Prototype products, applied demos, and the current 3D-photo toolchain.
↗
03
Code Repositories, model implementations, and public artifacts.
↗
Current tracks

Speech and audio

Hybrid recurrent-attention speech codec modeling, decoding experiments, and the next text-conditioned TTS branch under nani-speech-v0.x.x.

3D and rendering tools

Prototype work around portrait and volumetric capture, gaussian-splatting-adjacent tooling, and branded output workflows.

Notes

A stable entry point for independent technical work that spans research artifacts, prototypes, and public code without pretending those are the same thing.