Prompt Engineering
Learn how to write clear instructions, provide context, control output style, and improve AI results.
Learning Hub / AI Tools
English & Burmese
AI Tools, Prompt Engineering & Workflows
Learn how to use AI tools effectively, write better prompts, connect AI APIs, and design practical AI-powered workflows for content, business, development, and automation.
AI Workflow Pattern
Prompt to Action
User Input
Step 1
Prompt
Step 2
AI Model
Step 3
Safety Check
Step 4
Structured Output
Step 5
Workflow Action
Step 6
What You'll Learn
AI tools are more powerful when you understand how to guide them, connect them, and use them safely inside real workflows.
AI APIs
Understand how apps connect to AI models using secure server-side API calls.
AI Workflows
Combine prompts, tools, data, review steps, and automation into repeatable systems.
Safe AI Usage
Use AI without exposing secrets, private data, API keys, or sensitive business information.
Curriculum
AI ツール学習パス
順番に学ぶことも、必要なトピックへ直接進むこともできます。
基礎20 min
1. AI ツールとは
AI ツールとは何か、どのように回答を生成するか、実務でどこに役立つかを学びます。
AI assistantAI modelPromptContextOutputLimitationsHuman review
レッスンを開く基礎30 min
2. プロンプトエンジニアリングの基礎
より明確で有用、一貫した結果を得るためのプロンプト作成を学びます。
Clear taskContextRoleOutput formatExamplesConstraintsRevision prompts
レッスンを開く実践35 min
3. 実務で使うプロンプトパターン
文章、計画、コーディング、調査、デザイン、ビジネスワークフローで使える再利用可能なプロンプトパターンを学びます。
SummarizeRewriteCompareExplainGenerate ideasExtract dataCreate checklistTurn notes into action plan
レッスンを開く実践30 min
4. AI モデル選択
速度、コスト、推論、コーディング、文章作成、自動化に合わせた AI モデル選択を学びます。
Fast modelReasoning modelCoding modelMultimodal modelCost tradeoffLatencyQuality check
レッスンを開く開発者40 min
5. AI API のしくみ
アプリが API 経由で AI モデルへプロンプトを送り、構造化レスポンスを受け取る仕組みを学びます。
API requestAPI responseJSONHeadersServer-side callRate limitError handlingToken usage
レッスンを開く開発者35 min
6. 構造化 AI 出力
AI に JSON、表、チェックリスト、スキーマ、ワークフロー向け出力を返させる方法を学びます。
JSON outputSchemaValidationRequired fieldsParsingRetryFallback output
レッスンを開くAI ワークフロー45 min
7. AI ワークフロー設計
AI を一回限りのチャットではなく、大きなシステム内の一ステップとして設計する方法を学びます。
InputPrompt templateAI stepTool actionHuman approvalOutput destinationLoggingCost control
レッスンを開く応用35 min
8. コンテンツとマーケティング向け AI ツール
コンテンツ案、SNS 投稿、メール下書き、SEO アウトライン、動画台本、再利用に AI を使う方法を学びます。
Blog outlineSocial postEmailAd copyVideo scriptBrand voiceRepurposingReview checklist
レッスンを開く応用40 min
9. オートメーション向け AI ツール
要約、分類、翻訳、書き換え、抽出、ルーティングのためにワークフロー内で AI を使う方法を学びます。
AI summarizeAI classifyAI rewriteAI translateAI extractConditionActionHuman review
レッスンを開くセキュリティ40 min
10. AI 安全性と API キー保護
API キー、個人データ、機密性の高い業務内容を漏らさずに AI ツールを使う方法を学びます。
Server-side API callsSecret managementNo frontend keysNo keys in URLsRedacted logsRate limitsUsage limitsHuman approval
レッスンを開くPrompt Engineering Lab
Practice turning vague requests into clear, useful prompts. This lab is static and does not run AI calls.
Rewrite a Vague Prompt
Weak prompt
Write about AI.
Better prompt
Write a 700-word article for small business owners explaining how AI can save time in daily operations. Use simple language, practical examples, and a clear checklist at the end.
Specific task, audience, length, style, and output structure improve the result.
Add Context
Weak prompt
Make this better.
Better prompt
Rewrite this product description for a SaaS landing page. Keep it professional, clear, and conversion-focused. Do not change the main meaning.
AI needs context about the content, goal, and target format.
Control Output Format
Weak prompt
Give me ideas.
Better prompt
Give me 10 automation workflow ideas in a markdown table with columns: Use Case, Trigger, AI Step, Action, Business Value.
Output format reduces cleanup work.
Ask for Review
Better prompt
Review this workflow plan. Find missing steps, security risks, cost risks, and unclear user actions. Return the result as a checklist.
AI is useful for review, not only generation.
Create a Reusable Prompt Template
Better prompt
You are helping me create [content type] for [audience]. The goal is [goal]. Use [tone]. Include [required sections]. Avoid [things to avoid]. Return the result in [format].
Prompt templates help teams get consistent results.
AI API Concepts
API
A way for software systems to talk to each other.
Model
The AI engine that receives input and generates output.
Prompt
The instruction sent to the model.
System Instruction
A higher-level instruction that controls behavior, style, or rules.
Input Tokens
The amount of text or data sent into the model.
Output Tokens
The amount of text generated by the model.
Temperature
A setting that controls how creative or consistent the output may be.
JSON Mode / Structured Output
A way to ask the model to return predictable data.
Rate Limit
A limit on how many requests can be sent in a period of time.
Fallback Model
A backup model used when the first model fails or is too slow.
AI Gateway
A backend layer that protects API keys, tracks usage, and routes requests to model providers.
Security
Safe AI API Pattern
Keep provider keys and private data behind your backend or AI gateway. The public interface should never call AI providers directly with secret credentials.
Frontend UI
Your Backend / AI Gateway
AI Provider API
Backend validates result
Frontend displays safe output
Correct pattern: the browser talks to your backend, your backend protects keys and calls the AI provider, then the backend validates the result before returning safe output.
ThutaTech AI Tool Map
AI skills connect across writing, automation, app building, and visual design inside the ThutaTech Ecosystem.
Thuta Flow
Use AI steps inside workflows for summarizing, classifying, extracting, routing, and review.
ExploreThuta App Builder
Use AI to plan app ideas, shape user flows, and generate web or mobile project structure.
ExploreThuta Studio
Turn AI-assisted ideas into visual assets, documents, social graphics, and presentations.
ExploreReady to Use AI as a Real Workflow Skill?
Start with prompt fundamentals, then connect AI APIs, structured output, safety checks, and ThutaTech tools into repeatable systems.