Hi Product Hunters! 😻

👋 I am Harish, one of the founding team at voltaML. We are a bunch of deep learning engineers working on this product/library for close to a year now. We, as ML/DL engineers, are focused predominantly on model development, then we think about model deployment as a second thought and model optimization - it never crosses our brain. But guess what - 90% of AI costs is inference (says AWS).

⚙️ We started tinkering with model compilation and optimization and IT WAS HARD! There are a bunch of open-source compilers like TensorRT, Apache TVM, ONNX, TorchFX etc. to help optimize the models for a target hardware. But to go through their documentations and codes, building these compilers, setting them up in a particular environment, checking hardware and algorithm compatibility and finally making them work on YOUR MODEL, is a NIGHTMARE....seriously!

💪 So, to make your lives (and ours) easier, we have developed a library (which we will be open-sourcing soon) and a platform called voltaML. You can basically optimize and compile your ML/DL models to a specific target hardware and accelerate your inference throughput by up 10X - with a SINGLE LINE OF CODE. Or NO-CODE, if you prefer that. 2-3 clicks before deployment, and make your models go brrrr...🏎️!!!

🚀 While the platform is getting ready for a beta-release, we have created a demo app that you can try and see how stupid-easy it can be to optimize your model. Please do share it with your friends and colleagues and let us know how you like it.

🤖 AI is moving towards faster, cheaper, efficient and edge-compatible. With voltaML, our vision is to become the go-to community/platform for the best model optimization and compilation tools and algorithms.

⚡⚡ Try our voltaML demo platform today ⚡⚡

Cheers 🙌 🙌