jumplander/AIForge-1K-Architecture

AIForge-05-Architecture

Architecture Dataset for AI and Programming Tasks

Overview

AIForge-05-Architecture is a curated English dataset designed for AI systems working on architecture tasks in software engineering and programming.

Contents

• data.jsonl
• data.json
• metadata.json

Use Cases

• AI agent training
• Supervised fine-tuning
• Evaluation and benchmarking
• Software engineering research

Example Record

{
  "id": "AIARC_00001",
  "category": "architecture",
  "domain": "recommendations",
  "objective": "Design a monolith solution for a recommendations platform with low budget.",
  "requirements": [
    "Scale predictably",
    "Keep observability high",
    "Avoid unnecessary coupling",
    "Support future growth"
  ],
  "recommended_architecture": "A monolith approach with clear boundaries around the file and core business rules.",
  "components": [
    "API layer",
    "Domain service",
    "Persistence layer",
    "Async worker",
    "Monitoring and alerting"
  ],
  "tradeoffs": "Monolith improves the low budget story, but the team must keep ownership boundaries explicit.",
  "decision_rationale": "Choose the simplest architecture that satisfies the operational and product constraints.",
  "difficulty": 3
}

License

Released under CC BY 4.0.

Source

JumpLander