AI Tutorials
Hands-on AI tutorials written by AiDiscovery — from prompt engineering to building agents.
16 tutorials

Fine-tuning a Persian model with QLoRA on one GPU: step by step
With QLoRA and a single 24 GB GPU, adapt an open Persian-friendly model to your own tone and format: from dataset to adapter to deployment.
20 min
The complete LLM guide: RAG, fine-tuning and production evaluation
How to decide whether RAG is enough for your problem or fine-tuning is justified — and how to evaluate and operate an LLM system before and after launch.
25 min
AI agents in 2026: build, deploy and scale — the full guide
From the agent loop (reason → act → observe) to tools, memory, guardrails and deployment notes — a foundational guide to building your first real agent.
20 min
The realistic guide to mastering AI agents: a 6–9 month roadmap
A learning roadmap from programming foundations to deploying production agent systems — with a concrete project per stage and honest time estimates.
18 min
RAG vs. fine-tuning vs. AI agents: the decision guide
What each technique does, when it is justified and when it burns budget — comparing time, cost, risk and maintenance so you make an engineering decision, not a fashionable one.
12 min
How to make AI videos: from Midjourney to Sora and Veo 3
A two-step workflow that gives you control: build consistent still frames first, then add motion — with simple rules for consistency, movement and editing.
9 min
Prompt engineering best practices 2026: patterns and anti-patterns
Which patterns still work on current-generation models, which to drop, and how to test prompts like code.
12 min
The AI coding stack for 2026: every tool you need to build, ship and scale
Today’s stack in four layers — repo-aware agent, model choice, review and tests, shipping — plus the habits that actually speed you up and where projects fall apart.
14 min
Prompt engineering complete guide 2026: zero-shot to ReAct
Every major technique in order of complexity: zero-shot, few-shot, Chain-of-Thought, system prompts, RAG and ReAct — plus an honest list of what no longer works.
15 min
What is MCP (Model Context Protocol)? The 2026 guide
A simple, current explainer: the host/client/server architecture, the three capability types, and why MCP is called the "USB-C of AI".
10 min
Best AI coding tools: Cursor vs. Claude Code vs. Codex
Which one is for you? A practical comparison by workflow style, review appetite and budget — instead of chasing brand names.
10 min
10 good vs. bad ChatGPT prompts: what actually works in 2026
Before/after examples showing how context, roles, constraints and output format turn a useless answer into a precise, usable one.
8 min
Advanced MCP: from a toy server to production deployment
Topics the beginner tutorials skip: security and validation, sessions and state, tool design at scale, and keeping a server healthy in production.
1 h
Getting started with Midjourney: first image to mastering parameters
From your first prompt to the anatomy of a good one, aspect ratios, editing tools and building a personal prompt library — no prior background needed.
16 min
Build your first MCP server: a hands-on step-by-step guide
From zero to a working server: scaffold the project, define tools, connect a client and test — all in about an hour.
45 min
Making videos with Sora from scratch: a step-by-step guide
From your first generation to polished output: learning the interface, writing scene prompts, one-variable iteration and fixing the most common problems.
20 min