# ml-engineer

The ml-engineer subagent designs and runs the full ML model lifecycle, from data pipelines to deployment and monitoring. Use it for production MLOps work.

- subagent · `type=agent` · Claude Code, Codex, Cursor · ★ 1357 repo stars
- Author: qdhenry
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/qdhenry/Claude-Command-Suite/tree/HEAD/.claude/agents/data-ai/ml-engineer.md
- Updated: 2026-03-01
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/agents/qdhenry-ml-engineer

## What it does

ml-engineer helps plan an end-to-end machine learning system, covering data intake, feature pipelines, training, and model serving. It can work through deployment options such as TorchServe, TF Serving, and ONNX Runtime, with Docker and Kubernetes used for containerized delivery.

The subagent can shape CI/CD for model training and releases, add version control for datasets and models, and define checks for model quality, data drift, and concept drift. It also covers A/B tests, staged rollouts, shadow mode, and recovery to an earlier model version.

Its available tools include file editing, project search, shell commands, web research, and Task. The `context7` and `sequential-thinking` MCP integrations support library research and architecture analysis. If a project does not use Docker or Kubernetes, the container deployment path does not apply.

## Install

### Claude Code (native platform)

**Manually: put the files in place**
- File: `.claude/agents/ml-engineer.md` — For all projects use ~/.claude/agents/. Claude delegates by the description field or an @-mention

### Codex

**Manually: put the files in place**
- File: `.codex/agents/ml-engineer.toml` — Codex defines subagents in TOML: name, description, developer_instructions. Move the Markdown text into developer_instructions by hand. For all projects use ~/.codex/agents/

### Cursor

**Manually: put the files in place**
- File: `.cursor/agents/ml-engineer.md` — Cursor also reads .claude/agents/

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/qdhenry/Claude-Command-Suite/tree/HEAD/.claude/agents/data-ai/ml-engineer.md

## When to use it
- design an ML pipeline from data intake to serving
- build CI/CD for model training and deployment
- set up checks for model quality and data drift
- version datasets, features, and model artifacts
- plan an A/B test with a rollback path

## Example prompts
- Use the `ml-engineer` subagent to design an ML system from data intake through model serving.
- Ask `ml-engineer` to outline CI/CD that trains, validates, and gradually releases a model.
- Have `ml-engineer` define production monitoring for model quality, data drift, and rollback.

## Similar
- [mlops-engineer](https://affpapa.org/en/nevibecode/agents/voltagent-community-mlops-engineer.md) — subagent · 1009 installs
- [machine-learning-engineer](https://affpapa.org/en/nevibecode/agents/machine-learning-engineer.md) — subagent · 431 installs
- [senior-ml-engineer](https://affpapa.org/en/nevibecode/skills/senior-ml-engineer.md) — skill · 915 installs
- [ml-ops-engineer](https://affpapa.org/en/nevibecode/skills/ml-ops-engineer.md) — skill · 180 installs
- [dgx-spark-ops-engineer](https://affpapa.org/en/nevibecode/agents/dgx-spark-ops-engineer.md) — subagent · ★ 42780 repo stars
- [robotics-engineer](https://affpapa.org/en/nevibecode/agents/robotics-engineer.md) — subagent · ★ 2858 repo stars

## Other versions with the same name
- [ml-engineer · voltagent](https://affpapa.org/en/nevibecode/agents/voltagent-community-ml-engineer.md) — 1971 installs
- [ml-engineer · wshobson](https://affpapa.org/en/nevibecode/agents/ml-engineer.md) — 1971 installs
- [ml-engineer · voltagent](https://affpapa.org/en/nevibecode/agents/voltagent-ml-engineer.md) — ★ 6505 repo stars
- [ml-engineer · rohitg00](https://affpapa.org/en/nevibecode/agents/rohitg00-ml-engineer.md) — ★ 2858 repo stars

From the same set: [qdhenry/claude-command-suite](https://affpapa.org/en/nevibecode/by/qdhenry/claude-command-suite.md) — 112 cards

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
