Complete Aulafy map
Content index
Find a path, course, lesson, or article without navigating through menus. Search for a need or open the section that matters to you.
25Courses and lessons
201lessons
3Blog and updates
0Guides and quick entry points
237 visible items
Main sections
8Start from zeroA guided first task with no coding required.Learning pathsChoose a route by goal and experience.Course libraryBrowse every English course available on Aulafy.Full programmeSee the complete learning structure in one place.Continuous AI courseA structured programme from foundations to production.AI journalCurrent explainers, comparisons, and practical field notes.SearchSearch courses, lessons, and articles.Sources and review policyHow Aulafy verifies and updates its content.
Courses and lessons
25Understand what AI can do0 lessons
DeepSeek Harness (DSH), from zero to expert0 lessons
Aulafy foundations6 lessons
Codex for programmers10 lessons
- →Open course
- 01Install Codex and understand its surfaces
- 02Explore a repository without getting lost
- 03Request changes that can be verified
- 04Teach the repository with AGENTS.md
- 05Implement, test, and review the interface
- 06Work with Git and review changes
- 07Repository hygiene: remove redundancy without breaking behavior
- 08Control permissions, network, and secrets
- 09Automate tasks with codex exec
- 10Final project: deliver a complete change
AI Router and content system6 lessons
Claude Code, from zero to pro24 lessons
- →Open course
- 01Installation
- 02First Steps
- 03CLI, app, and mobile
- 04Practical Recipes
- 05Guided Projects
- 06Writing good prompts for Claude Code
- 07Control Context and Costs
- 08Glossary
- 09SMBs and Office Work
- 10Technical playbooks
- 11Skills
- 12Subagents
- 13Plugins
- 14Pro Workflows
- 15Commands
- 16Configuration
- 17MCP Servers
- 18Hooks
- 19Permissions
- 20Advanced Usage
- 21Frequently Asked Questions
- 22Troubleshooting
- 23Resources
- 24Comparison
Claude Code + Local AI36 lessons
- →Open course
- 01The Terminal Without Fear (CLI)
- 02How to Work with Your Projects
- 03Writing good prompts for local AI
- 04Local AI: choose your model
- 05Ollama from Scratch: Install & Run Local AI in 2026
- 06LM Studio from Scratch: Install and Run Local AI
- 07GGUF Quantization: Q4, Q5, and Q8
- 08Connect Claude Code to Your Local AI
- 09Fix Ollama Errors
- 10Ollama Caps Context at 4K or 32K? Check What It Actually Loaded
- 11Ollama Not Using GPU on Windows? NVIDIA, AMD & WSL2 Fixes
- 12Debug and protect your work
- 13Minimum Hardware for Local AI in 2026
- 14Ollama, vLLM, SGLang, and MLX: What to Use
- 15Windows/WSL2 vs Mac M4 for Local AI
- 16AI Homelab with Used RTX 3090s
- 17Quantization and models for coding
- 18Local code agents with Ollama
- 19Local coding agent with Hermes and Ollama
- 20Advanced Hermes with Gemma and MLX
- 21Tool calling with local models
- 22Open WebUI + Ollama + Qdrant
- 23Open WebUI troubleshooting and lightweight production
- 24Chatbot That Cites the Law (RAG)
- 25Lexia-style web app with RAG
- 26Ask Your PDFs
- 27Chatbot that listens and speaks
- 28Text to Audio
- 293D Simulations
- 30Speaking Avatar
- 31WordPress Theme with AI
- 32Website for your service
- 33Assistant for freelancers
- 34Study App
- 35Publish your app on the internet
- 36Home cluster with exo
Generative AI: image, voice and video7 lessons
Games, 3D and simulations with AI6 lessons
AI security and evaluation7 lessons
Local MLOps and model deployment10 lessons
- →Open course
- 01Serving map: Ollama, llama.cpp, vLLM
- 02VRAM and OOM: vLLM vs Ollama/llama.cpp
- 03llama.cpp server locally
- 04vLLM with OpenAI-compatible API
- 05LiteLLM as a Gateway and Cost Control
- 06Hybrid local/cloud routing with LiteLLM
- 07Migrate a cloud agent to local
- 08Observability with Langfuse and OpenTelemetry
- 09Queues, rate limits, and cache
- 10Project: private mini LLM platform
Fine-tuning and post-training for LLMs8 lessons
Agents and automation20 lessons
- →Open course
- 01A real map of agents in 2026
- 02From chat to agent: model calls, tools, MCP, RAG, and loops
- 03Subagents with Roles and Limits
- 04Building an R-style agent CLI
- 05Hooks: Deterministic Automation
- 06Secure and auditable skills
- 07MCP without giving away your keys
- 08MCP for Local Tools
- 09Create a Custom MCP Server
- 10GitHub Actions and Routines
- 11Project: 24/7 agent with inbox
- 1224/7 offline agents on real hardware
- 13OOM and memory management
- 14Retries, idempotency, and exactly-once
- 15Persistent State and Crash Recovery
- 16Infinite Loops and Cost Control
- 17Local Multi-Agents, Memory, and Loops
- 18Shared Persistent Memory
- 19MCP Verification Before Merge
- 20MCP Governance and Reliable Knowledge
Production agents with LangGraph and n8n9 lessons
- →Open course
- 01LangGraph, n8n, and CrewAI: What to Use
- 02LangGraph vs CrewAI vs n8n in 2026
- 03State, memory, and controlled loops
- 04n8n as a tools layer
- 05Human approvals and permissions
- 06Error Recovery and 503 Failures
- 07Evals, logs, and observability
- 08Observability for local agents
- 09Project: inbox agent for SMBs
Self-hosted AI automation for small businesses10 lessons
- →Open course
- 01Stack map: n8n, Open WebUI, and models
- 02Cheap Docker and VPS without the mess
- 03Docker Compose for a Complete Local AI Stack
- 04n8n on Windows: Docker, WSL, and Common Errors
- 05Ollama and Open WebUI as a private interface
- 06Private RAG with n8n, Qdrant, and Ollama
- 07n8n with webhooks and credentials
- 08Human approvals before acting
- 09Queues, Backups, and Monitoring
- 10Project: internal support for an SMB
Advanced and secure RAG14 lessons
- →Open course
- 01Useful RAG: much more than PDF chat
- 02Ingestion, cleaning, and chunking
- 03Advanced Chunking for Private RAG
- 04Local GraphRAG and graph-based memory
- 05OCR and Tables in Real-World PDFs
- 06Embeddings and Vector Databases
- 07Hybrid search and reranking
- 08Qdrant Multi-User and Permissions
- 09Local second brain with Obsidian and RAG
- 10Prompt injection in RAG
- 11Debugging RAG: grounding and the full prompt
- 12Structured responses with citations
- 13RAG Evals with Metrics
- 14Evals, citations, and traceability
AI for small businesses and freelancers20 lessons
- →Open course
- 01Useful AI Map for a Small Business
- 02AI literacy for SMEs: inventory and internal policy
- 03AI Act transparency: audit chatbots and synthetic content
- 04Diagnosis and first AI pilot for an SME
- 05Ollama for an SME: a local, secure, and measurable trial
- 06GDPR basics for using AI without surprises
- 07From pilot to reliable flow: schema, sources, abstention and validation
- 08Failure testing and metrics before automating
- 09Minimum operations: logs, owner, backups and degraded mode
- 10Permissions and agents: automate only when the flow is ready
- 11Emails: Classify and Create Drafts
- 12Invoices: extract data and review
- 13Invoices, OCR, and Verifactu with Human Review
- 14Budgets, Excel, and Sheets
- 15Convert a bank statement to CSV without inventing transactions
- 16Minimal local RAG with SQLite and Ollama
- 17First secure MCP for the office: local and read-only
- 18MCP 2026-07-28: safe migration and read-only server
- 19WhatsApp and Telegram with Human Approval
- 20AEO for SMEs: appearing in AI answers
AI Consultant: From Assessment to Production8 lessons
- →Open course
- 01What an AI consultant actually does
- 02Consultant, developer, or automation agency?
- 03The AI consulting lifecycle
- 04Build your AI Consultant Operating Model
- 05Read a business before proposing AI
- 06Revenue, customers, and operating constraints
- 07Knowledge flows, handoffs, and information loss
- 08SIPOC and value-stream mapping
Choose your first useful task0 lessons
Write a clear request0 lessons
Check before you trust0 lessons
Protect your information0 lessons
Improve a draft without losing facts0 lessons
Use AI to support your learning0 lessons
Complete your first AI project0 lessons
Blog and updates
3Stop Comparing AI Models by Token Price: Why Cost per Task Matters MoreToken prices can mislead when comparing AI agents. Learn to calculate model cost, retries, human review and cost per accepted task.GPT-6 Astra: What Can It Actually Do? Real Use Cases, Benchmarks and How to Use ItWhat can GPT-6 Astra actually do? Real computer-use, coding, research and business examples, benchmarks, prompting tips and current availability.The latest local AI models in 2026: Kimi K3, Inkling, Laguna, DeepSeek, Qwen, Muse and NemotronAn updated comparison of the newest open-weight models released in July and August 2026, including hardware, benchmarks and realistic local use.