Hi, I'm Suraj 👋

Principal Software Engineer at Microsoft, working on AI/ML infrastructure and Kubernetes. CKA & CKS certified · Conference speaker · Open source contributor

---

📝 Latest Blog Posts Suraj Deshmukh's talks at conferences Ghostty as a Shell for Herdr Interactive Visualizations for the book 'Inference Engineering' The Ironies of AI Coding: Combating Cognitive Debt with Visual PRs Will AI Agents be the Death of YAML?

🔨 Recent Contributions surajssd/llm-k8s — 1 day ago surajssd/blog_contents — 1 day ago surajssd/surajssd.github.io — 1 day ago surajssd/dotfiles — 1 day ago surajssd/libnotion — 1 week ago

🔀 Recent Pull Requests

• fix(metrics): handle missing CPU sequence lengths on sgl-project/sglang — 1 week ago
• feat(vllm-tensorizer): Bump DeepGEMM to a6b593d2 for SM12x attention kernels on coreweave/ml-containers — 1 month ago
• fix(renderer): honor tool_choice=none and full reasoning_effort range in DeepSeek native formatters on ai-dynamo/frontend-crates — 1 month ago
• fix: DeepSeek-V4 serving fixes, streaming 4xx, deployment default_thinking on ai-dynamo/dynamo — 1 month ago
• fix(test): stop huggingface.test reloading the shared module singleton on ai-runway/airunway — 1 month ago