Skip to content
Agentic Levels
Start hereLevelsTracksPromptsAbout
Get my levelGet my level

  • Start hereYour first useful task with AI.
  • Level quiz8 questions. No AI needed.
  • Fluency ScoreYour Level and a score, from your own AI.

  • LevelsThe 11 levels, in order.
  • LessonsShort lessons, filter by level.
  • TracksGuided paths by tool, role or goal.
  • Set up your AI tool
  • Starter packFree files that help AI remember.

  • PromptsReady prompts to copy and adapt.
  • ChatGPT vs Claude vs GeminiPick the right one for the job.
  • How-tos
  • ReferenceCommands for Claude Code, Codex, Antigravity.
  • ResourcesHand-picked reading and tools.

  • Your progress
  • Your next moveAn interview that picks your next skill.
  • Your next lessonsThe next lessons at your level.
  • Updates

  • Pricing
  • About
  • FAQ & Feedback
  • Work with us
  • How we handle your data
Sign inPreferences
AboutPricingHow-tosPromptsStarter packMy stackHow we handle your data

© 2026 Fuentes Studio

PrivacyTermsRefundsAccessibilityAI content policyCredits
yourCouncil
Ready to help
✦

What do you want to understand?

Ask anything about what you're learning.

←How-to guides
  1. Resources
  2. ›How-to guides

How to use AI Engineering Hub

93+ production-ready reference implementations for AI engineering patterns: RAG, agents, multimodal, voice, and fine-tuning with beginner-to-advanced roadmap.

Levels 3 to 5On GitHub

When you need to add RAG, an eval harness, an agent loop, or an MCP server to your project, starting from scratch is the slowest path. AI Engineering Hub is a library of 93+ complete, runnable examples organized by pattern. Find the one closest to your use case and work from it. Categories span LLMs, RAG, AI agents, multimodal systems, voice agents, fine-tuning, and enterprise-grade agent systems.

Before you start

Python 3.10+ and pip. Specific examples may need additional API keys. You can browse without any setup.

  1. 1Browse by pattern category
  2. 2Clone and run the example closest to your use case
  3. 3Ask Claude to explain and adapt it to your stack
  4. 4Move the adapted code into your project and verify on real input

Steps

01

Browse by pattern category

The repo is organized into top-level folders: RAG, Agents, Evals, MCP, Fine-tuning, and more. Open the folder that matches what you are building. Skim the README in that folder before opening any code. The README tells you which example is the simplest entry point.

02

Clone and run the example closest to your use case

Install that example's requirements and run it on your own data, not the demo data. A reference implementation running on your actual input is more useful than one running on a pre-packaged example.

Run the example
git clone https://github.com/patchy631/ai-engineering-hub cd ai-engineering-hub/[PATTERN_FOLDER]/[EXAMPLE_FOLDER] pip install -r requirements.txt [RUN_COMMAND]
[PATTERN_FOLDER]Top-level pattern folder e.g. rag
[EXAMPLE_FOLDER]Specific example folder e.g. basic-rag-pipeline
[RUN_COMMAND]Command to run the example e.g. python main.py
03

Ask Claude to explain and adapt it to your stack

Open the example's main file in Claude Code and ask Claude to walk through what each section does, then adapt the parts that do not match your stack. The goal is to own the code, not copy-paste it.

Ask Claude this
I'm looking at this reference from ai-engineering-hub: [EXAMPLE_FILE_PATH] My stack is: [YOUR_STACK] 1. Walk me through what each section does in plain language. 2. Identify the parts that need to change for my stack. 3. Rewrite those parts and explain each change.
[EXAMPLE_FILE_PATH]Path to the example file e.g. ai-engineering-hub/rag/basic-rag-pipeline/main.py
[YOUR_STACK]Your current tech stack e.g. Next.js frontend, Supabase for storage and vector search, Anthropic SDK
04

Move the adapted code into your project and verify on real input

Wire the adapted code to your actual data source and run it against a real query. Most errors at this stage are missing environment variables or a library version mismatch. Read the error carefully before debugging.

How do you know it worked?

  • You ran the reference example on your own data and it produced output.
  • You can explain what each major section of the adapted code does without looking it up.
  • The adapted code is in your project and wired to your actual data source.
  • A query you care about returns a sensible result.

If something looks off

Version-pinned examples vs. newer library versions
Reference implementations freeze dependencies at writing time. Check requirements.txt against current library versions before running. Outdated versions often have breaking API changes.
Testing on demo data only
Demo data is chosen to make examples look good. Your data has noise and edge cases. Always test on real inputs before calling the integration complete.
Copying the entire example instead of the pattern
Hub code is reference material, not production code. It lacks error handling, logging, and rate-limit management. Extract the pattern, then add production concerns in your own codebase.

Skills & Commands Reference

The major pattern categories in the hub and when to reach for each.

Command / SkillWhat it doesWhen to use it
RAG pipelinesChunk → embed → retrieve → generate reference implementations for document Q&AAdding question-answering over private documents or knowledge bases
Agentic loopsLLM-driven planning and tool-use loops with self-correctionTasks requiring multi-step reasoning, tool calls, or iterative refinement
Structured outputsPatterns for forcing JSON or typed responses via tool-call or response_formatDownstream code needs to parse the model's output reliably
Eval harnessAutomated testing of LLM outputs against ground-truth or rubricsShipping a model or prompt change and need to verify no regression
MCP examplesadvancedModel Context Protocol server reference implementationsBuilding a custom MCP server and need a working starting point
Multi-agent patternsadvancedOrchestrator → worker setups with shared context and message passingTasks too large for a single context window or that parallelize naturally

What you'll do, 4 steps

  1. 1Browse by pattern category
  2. 2Clone and run the example closest to your use case
  3. 3Ask Claude to explain and adapt it to your stack
  4. 4Move the adapted code into your project and verify on real input

Was this helpful?