Skip to content
View ntny's full-sized avatar
1 follower · 1 following

Achievements

Achievement: Pull Sharkx2Achievement: QuickdrawAchievement: Pair ExtraordinaireAchievement: Arctic Code Vault Contributor

Achievements

Achievement: Pull Sharkx2Achievement: QuickdrawAchievement: Pair ExtraordinaireAchievement: Arctic Code Vault Contributor

Block or report ntny

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
ntny/README.md

Hi there 👋

Anton Pechenin

AI Infrastructure Engineer focused on inference infrastructure, ML platform runtime, and workflow orchestration.

Kubeflow member and OSS contributor. I work on backend/runtime architecture, reliability, execution semantics, and performance for ML workflow systems.

OSS

Kubeflow Pipelines

Selected merged work:

  • #12023 - central driver architecture proposal based on Argo Workflows Agent
  • #12010 - API server gRPC metrics and execution spec reporting optimization
  • #12610 - recurring runs queue throughput optimization
  • #12648 - reconciliation bug fix
  • #11673 - execution-level retry fix for Argo backend
  • #11585 - run retry fix for Argo
  • #11925 - launcher executor input parameter fix

Argo Workflows

  • #16075 - WorkflowTaskSets size reduction for large workflows
  • #15440 - Configure the executor plugin at the workflow level

NVIDIA Dynamo / AIPerf

  • #994 - Socket buffer fallback for macOS compatibility
  • #1153 - Support SGLang Speculative Decoding Metrics

In Progress SGLang

  • #31193 - Prevents SGLang server failure mode caused by unbounded Prometheus metric cardinality

Publications

Focus

  • LLM inference systems
  • Inference runtime optimization
  • GPU performance
  • Distributed runtime reliability
  • Kubernetes-native workflow systems
  • Backend architecture

Pinned Loading

  1. pipelines pipelines Public

    Forked from kubeflow/pipelines

    Machine Learning Pipelines for Kubeflow

    Python

  2. argo-workflows argo-workflows Public

    Forked from argoproj/argo-workflows

    Workflow Engine for Kubernetes

    Go

  3. argo-executor-plugin-demo argo-executor-plugin-demo Public

    Python

  4. aiperf aiperf Public

    Forked from ai-dynamo/aiperf

    AIPerf is a comprehensive benchmarking tool that measures the performance of generative AI models served by your preferred inference solution.

    Python

  5. vllm vllm Public

    Forked from vllm-project/vllm

    A high-throughput and memory-efficient inference and serving engine for LLMs

    Python