Description du poste
About KogAt Kog, we're building real-time LLM inference for AI agents.
Coding agents execute sequences of tasks where each step depends on the model's previous response. The faster the model generates tokens, the faster these agents can reason, code, and act.
Our approach is to co-design model architecture, the inference engine, and low-level GPU kernels to minimize generation latency on standard NVIDIA and AMD datacenter GPUs.
We've developed several technologies to make this possible:
Monokernel: a persistent GPU program that keeps the entire token generation loop on the GPU, avoiding CPU handoffs.
KCCL: our custom inter-GPU communication layer, designed to minimize latency.
Delayed Tensor Parallelism (DTP): a model architecture that overlaps communication with computation and weight loading.
Laneformer 2B: our code-generation model demonstrating this co-design approach.
Our engineers work directly with GPU hardware, using CUDA, PTX, HIP, and CDNA ISA to optimize execution.
Implementing and optimizing a new AI model for a specific GPU architecture can take months of manual engineering. Meanwhile, new models are constantly being released.
Our goal is to use AI agents to implement and optimize the models we choose on the GPUs we choose, in a fraction of the time required today.
These agents will build on the GPU optimizations and hardware insights our team has developed. They'll explore optimization strategies, generate and modify kernels, verify correctness, benchmark perfor