Need to build prototype with RT Audio processing & efficient ML TOPS

[kparasha]

I am working on a hardware prototype that requires Real-time audio processing (DSP) & ~3-ish TOPS on an Int8 NPU on board to run some AI models and preferably even intent processing with like tiny-tiny LLMs. Has someone worked on sth like this here? I am on a tight budget and timeline to get into a hardware incubator in SF 💯