MLOps Engineer

Automating the ML model lifecycle: from training to production monitoring

ML & AI Engineering Junior Middle Senior Lead / Staff Principal
完整矩阵 职业路径 PDF
56 技能
5 级别
123 必备
280 要求

MLOps Engineer是ML & AI Engineering族群中的角色。涵盖5个级别的56项技能(从Junior到Principal)。其中123项为必备技能。关键领域:Programming Fundamentals, Backend Development, Database Management。

技术栈

Junior Python, MLflow, DVC, Docker, Airflow basics, Kubernetes basics, Git, Prometheus basics
Middle Python, Kubeflow/MLflow, Docker/Kubernetes, Airflow, Feature Store (Feast), Seldon/BentoML, Terraform, GitHub Actions
Senior Kubeflow/Vertex AI, Triton Inference Server, ONNX/TensorRT, Kubernetes (GPU scheduling), Ray Serve, Custom operators, Prometheus/Grafana
Lead / Staff ML Platform architecture, Multi-model serving, GPU cluster optimization, A/B testing infra, Model governance
Principal Enterprise MLOps, Multi-cloud ML, LLM infrastructure (vLLM, TGI), Cost optimization at scale

各级别重点

Junior

Setting up ML pipelines. Working with MLflow/DVC. Containerizing models. Monitoring inference. Automating routine tasks.

Middle

Designing CI/CD for ML. Setting up feature store. Automating training and deployment. Drift monitoring. GPU cluster management.

Senior

MLOps platform architecture. Inference optimization (Triton, ONNX). Real-time serving. GPU autoscaling. Cost optimization.

Lead / Staff

MLOps platform strategy. ML lifecycle standards. Coordination with ML and backend teams. Vendor evaluation.

Principal

Enterprise MLOps architecture. Multi-cloud ML infrastructure. LLM deployment strategy. Industry best practices.

技能矩阵

56 技能 × 5 级别. 点击单元格查看详情。

A Awareness W Working V Advanced E Expert

AI-Assisted Development

4 技能
技能 Jun Mid Sen Lead Princ
GitHub Copilot A W A E E
Cursor IDE A W A E E
ChatGPT / Claude A W A E E
Prompt Engineering for Code A W A E E

API & Integration

4 技能
技能 Jun Mid Sen Lead Princ
REST API Design A W A E E
GraphQL Design A W A E E
gRPC & Protocol Buffers A W A E E
API Documentation A W A E E

Architecture & System Design

1 技能
技能 Jun Mid Sen Lead Princ
System Design Fundamentals A W A E E

Backend Development

5 技能
技能 Jun Mid Sen Lead Princ
Python Web Frameworks A W A E E
Apache Kafka A W A E E
Redis A W A E E
Task Queues A W A E E
S3 / Object Storage A W A E E

Cloud & Infrastructure

8 技能
技能 Jun Mid Sen Lead Princ
Docker A W A E E
Container Security Scanning A W A E E
Kubernetes Core A W A E E
Kubernetes Advanced A W A E E
Helm A W A E E
Terraform A W A E E
AWS A W A E E
Network Fundamentals A W A E E

Database Management

3 技能
技能 Jun Mid Sen Lead Princ
PostgreSQL A W A E E
Database Indexing A W A E E
Query Optimization A W A E E

DevOps & CI/CD

3 技能
技能 Jun Mid Sen Lead Princ
GitHub Actions / GitLab CI A W A E E
GitLab CI/CD Advanced A W A E E
ArgoCD A W A E E

Machine Learning & AI

6 技能
技能 Jun Mid Sen Lead Princ
MLflow A W A E E
Feature Stores A W A E E
Model Serving A W A E E
Experiment Tracking A W A E E
ML Pipelines A W A E E
Model Monitoring A W A E E

Observability & Monitoring

5 技能
技能 Jun Mid Sen Lead Princ
Structured Logging A W A E E
Prometheus & Grafana A W A E E
Custom Business Metrics A W A E E
OpenTelemetry A W A E E
SLI / SLO / SLA A W A E E

Programming Fundamentals

8 技能
技能 Jun Mid Sen Lead Princ
Algorithms & Complexity A W A E E
Data Structures A W A E E
OOP & SOLID Principles A W A E E
Design Patterns A W A E E
Multithreading A W A E E
Async Programming A W A E E
Code Quality & Refactoring A W A E E
Type Safety & Type Systems A W A E E

Security

3 技能
技能 Jun Mid Sen Lead Princ
OWASP & Application Security A W A E E
Secure Coding Practices A W A E E
JWT / OAuth2 / OIDC A W A E E

Testing & QA

4 技能
技能 Jun Mid Sen Lead Princ
Unit Testing A W A E E
Integration Testing A W A E E
E2E Testing A W A E E
Load Testing A W A E E

Version Control & Collaboration

2 技能
技能 Jun Mid Sen Lead Princ
Git Advanced A W A E E
Code Review A W A E E

常见问题

MLOps Engineer角色需要哪些技能?

MLOps Engineer角色需要56项技能,其中123项为必备。技能分布在5个级别:从Junior到Principal。 查看完整矩阵.

如何在MLOps Engineer角色中晋升到下一级别?

使用等级计算器评估您当前的级别并获取个性化建议。系统将显示晋升所需发展的技能。

MLOps Engineer角色使用什么技术栈?

技术栈包含5种不同级别的技术。 Python, MLflow, DVC, Docker, Airflow basics, Kubernetes basics, Git, Prometheus basics, Python, Kubeflow/MLflow, Docker/Kubernetes, Airflow, Feature Store (Feast), Seldon/BentoML, Terraform, GitHub Actions, Kubeflow/Vertex AI, Triton Inference Server, ONNX/TensorRT, Kubernetes (GPU scheduling), Ray Serve, Custom operators, Prometheus/Grafana...

社区如何定义MLOps Engineer角色的要求?

角色要求由社区通过提案系统制定。任何成员都可以提出修改建议,经过投票和专家评审后生效。

社区

👁 关注 ✏️ 建议修改 登录以建议修改
📋 提案
暂无提案 MLOps Engineer
正在加载评论...