The UnFiltered Media
GPU Optimization, Automated
RightNow AI just open-sourced AutoKernel — an AI agent that optimizes GPU kernels while you sleep.
RightNow AI has released AutoKernel, an open-source framework that uses an autonomous LLM agent loop to automatically optimize GPU kernels…
The UnFiltered Media
What RightNow AI Built
AutoKernel wraps an LLM inside an iterative agent loop. It generates, benchmarks, and refines Triton kernels for any PyTorch model automatically.
The UnFiltered Media
No CUDA Expertise Needed
RightNow designed AutoKernel to eliminate the need for specialized kernel engineers. Set it running overnight and wake up to faster code.
The UnFiltered Media
Works With Any Model
Custom transformers, diffusion backbones, fine-tuned LLaMA variants — AutoKernel handles arbitrary PyTorch architectures out of the box.
The UnFiltered Media
Fully Open Source
RightNow AI made a strategic choice to open-source the entire framework. Community adoption and contribution are core to the project's mission.
Read the Full Article
Discover how RightNow AI's AutoKernel works under the hood and what it means for the future of GPU programming.