DevOps / SRE engineer β Kubernetes, Linux internals, C/C++ & Python. I like the boundary between infrastructure and low-level code, and I contribute fixes upstream when tools break under me.
conan-io/conan β the C/C++ package manager (9.4kβ )
- β
merged #19965 β fix
cppstd_flagfor theclang-cldriver in C++23 / C++26 /gnu++modes - π in review #19963 β
conan.tools.gnu.is_mingw()helper (+ docs PR) - π¬ RFC #19964 β offline sources fallback for air-gapped builds
ollama/ollama β local LLM runtime (176kβ )
- π in review #16661 β mlxrunner: validate affine quant params against packed tensor shapes
- π in review #16663 β convert: read top-level text fields for text-only gemma4 configs
- DevOps / SRE engineer with strong background in C/C++ and Python
- Deep understanding of Linux internals, networking, processes and memory
- Experience operating and debugging production systems
- Run a personal Proxmox + Kubernetes homelab, fully managed with Terraform + Ansible (IaC, no snowflakes)
- Build offline-first AI infrastructure: local LLM serving, WhisperX/pyannote transcription pipeline, RAG on Qdrant
- Prefer automation, observability and reliability over magic
- Kubernetes internals & networking
- Local / offline-first AI infrastructure (MLX, llama.cpp, Ollama)
- High-load and fault-tolerant systems
- Observability: metrics, logs, tracing
- Performance profiling & low-level Linux debugging (strace, perf, gdb)
- Infrastructure automation
- Algorithms and data structures
- System-level thinking
- Focus on correctness and performance
π LeetCode profile



