I build kernel-leaning systems for AI infrastructure, KV-cache orchestration, memory hierarchy control, and low-latency runtime behavior.

Most of my work sits close to the machine: Linux control planes, kernel-facing memory and I/O experiments, CPU scheduling and latency behavior, KV-state movement, and research prototypes that make systems ideas concrete.

• I like working on the hard parts of systems for AI: memory placement, I/O paths, scheduler behavior, latency control, and observability.
• I build across the stack, but I naturally gravitate toward Linux, kernel-adjacent interfaces, CPU and memory behavior, and runtime control planes.
• I care about making low-level work legible, so I pair code with diagrams, RFC-style docs, and architecture-driven writeups.

• KV-cache orchestration and memory residency control
• Linux kernel control planes for inference workloads
• CPU, IRQ, scheduler, and latency-path tuning
• Storage and I/O behavior for decode-critical serving paths
• Systems observability for real AI infrastructure

• kairo-io
  AI KV-cache-aware Linux block I/O work focused on decode-priority scheduling, placement metadata, NVMe backend mapping, and kernel tracepoint visibility.

• kv-cpu-driver
  A Linux control-plane and hardware-interface scaffold for semantic KV-cache orchestration, with CPU-facing coordination, RTL structure, and FPGA emulation paths.

• linux-hbf-control-plane
  An RFC-style Linux exploration of runtime-guided memory placement, prefetch, promotion, and demotion for future CXL/HBF-era inference systems.

• kernel-dvfs-agentic-latency
  A kernel latency control-plane project spanning DVFS, cpuidle, IRQs, scheduler behavior, MM, VFS, I/O, and cgroup budget control for latency-sensitive AI execution.

• ai-host-observability
  Linux-first host observability for GPU and RDMA systems, built to surface memory pressure, reclaim, PCIe, NUMA, IRQ, and host-side failure signals before they become incidents.

• Portfolio: manishklach.github.io
• Repositories: github.com/manishklach

I regularly publish architecture-driven essays and technical companion material around Linux systems, AI infrastructure, memory-centric design, and runtime control.

• Kernel and kernel-adjacent experiments
• KV-cache, memory, CPU, and I/O control-plane ideas
• Systems observability and performance tooling
• RFC-style writeups, architecture notes, and technical docs
• Research prototypes tied to real implementation artifacts

If you're into Linux internals, memory systems, KV-cache control, CPU latency paths, or kernel-facing AI infrastructure work, you'll probably find something interesting here.