design

ai-performance-engineering: Boost AI Model Speed & Scale

Name: ai-performance-engineering: Boost AI Model Speed & Scale
Author: cfregly

Offers Python code, labs, and resources to optimize GPU use, run distributed training, and scale inference—perfect for founders building high‑performance AI systems. 2.1k+ GitHub stars

From the repo · “Code, labs, and resources for O'Reilly AI Systems Performance Engineering: GPU optimization, distributed training, inference scaling, and full-stack tuning.”

2,090 stars+29★ this week294 forksPythonGuide quality 8/10 · reviewed Oct 22Updated 10/22/2026100% free · open source

What it does

The ai-performance-engineering tool provides resources and labs for optimizing AI system performance, including GPU optimization, distributed training, and full-stack tuning, specifically for founders

When to use it

•When you need to optimize the performance of your AI models for production
•When you're experiencing bottlenecks in your AI pipeline and need to scale
•When you want to improve the efficiency of your AI system's GPU usage

Ready-to-paste prompt

Run `python lab-02-distributed-training/lab.py --num-gpus 4` to start the distributed training lab with 4 GPUs

Heads up: Make sure you have the necessary GPU drivers and CUDA version installed, as specified in the README, to run the GPU optimization labs

Saves to your device

Use with Claude

New

Skip the builder — one click puts this in Claude, Cursor, Antigravity and more.

✅ Light setup

Installs with a command or two; your AI agent can do it for you.

Try it instantly — no install

Claude Code

mkdir -p ~/.claude/skills/ai-performance-engineering && curl -fsSL https://workflowstacks.com/api/skills/ai-performance-engineering/claude-skill -o ~/.claude/skills/ai-performance-engineering/SKILL.md

Open in another AI app

Opens the app with this repo with the prompt ready to go — no copy-paste needed.

Connect the whole catalog (MCP)

claude mcp add --transport http workflowstacks https://workflowstacks.com/api/mcp

Adds a WorkflowStacks connector to Claude Code: search and load any skill here by chatting.

What's inside — free to inspect

No purchase needed

Read the entire source before you build — unlike paid marketplaces that hide it behind a buy button.

top-level files

folders

2773.9M

source size

Apache-2.0

license

Key files

AGENTS.md

README.md

File tree

.agents/

.claude/

.github/

code/

docs/

img/

meetups/

resources/

.env.sample

.gitignore

.gitmodules

AGENTS.md

AUDIT_REMEDIATION_PLAN.md

CONTRIBUTING.md

HANDOFF.md

LICENSE

README.md

Use with Claude

New

Skip the builder — one click puts this in Claude, Cursor, Antigravity and more.

✅ Light setup

Installs with a command or two; your AI agent can do it for you.

Try it instantly — no install

Claude Code

mkdir -p ~/.claude/skills/ai-performance-engineering && curl -fsSL https://workflowstacks.com/api/skills/ai-performance-engineering/claude-skill -o ~/.claude/skills/ai-performance-engineering/SKILL.md

Open in another AI app

Opens the app with this repo with the prompt ready to go — no copy-paste needed.

Connect the whole catalog (MCP)

claude mcp add --transport http workflowstacks https://workflowstacks.com/api/mcp

Adds a WorkflowStacks connector to Claude Code: search and load any skill here by chatting.

Quick Actions

Details

Creator

cfregly

Language

Python