Use Case Taxonomy
A systematic classification of generative AI patterns for software development and content creation
Basic Patterns
Foundational generation patterns: creation, transformation, and synthesis
Advanced Patterns
Incremental, hierarchical, and iterative generation strategies
Combined Workflows
Real-world pipelines combining multiple patterns
Basic Generative Use Cases
Content Creation
FoundationalGenerate new content from scratch based on descriptions or prompts.
"Create a REST API for user management"
UserController.kt
Examples
- Generate file from description
- Generate images from text prompts
- Create boilerplate code from specifications
- Generate documentation from requirements
Content Transformation
TransformTransform existing content while preserving its essential meaning or structure.
Original Content
Transformed Content
Examples
- Summarization of articles
- Translation between languages
- Edit / feature addition to existing code files
- Code refactoring and optimization
Synthesis
CombineCombine multiple inputs to create unified, coherent output.
Examples
- Synthesis by Example: Generate new content based on multiple input examples
- Pattern Extraction: Identify common patterns from multiple inputs
- Mass Summarization: Summarize multiple documents into a single cohesive summary
Analogical Generation
AdvancedCreate new content by combining and modifying existing pieces through analogy.
Concept
Transfer features from one class/file to another — a sort of "transgenic" content generation.
UserService (A) - BasicCRUD (B) + CachingLayer (C) → CachedUserService (D)
Advanced Generative Use Cases
Incremental Generation
ExtendGenerate additions to existing content, continuing from where it left off.
A
A + δ
💡 Key Insight
Essentially telling the model to "continue" from where it left off, useful for long-form content generation.
Hierarchical Generation
ScaleGenerate complex content through multiple stages of refinement and transformation.
💡 Key Insight
Allows for exponential growth in content volume. LLM output is generally constrained by input size, but hierarchical generation bypasses this limitation.
Applications
- Spec-driven development
- Large codebase generation
- Documentation hierarchies
Iterative Editing
RefineRepeatedly refine content through multiple editing passes.
⚠️ Considerations
- Patching Strategy: Full replacement vs. modification of existing content
- Convergence: Monitor for stability and diminishing returns
- Quality Gates: Define stopping criteria
Error Correction
FixIdentify and correct errors using a feedback loop with external validation.
Tests / Linter
⚠️ Considerations
- May require many iterations to achieve desired quality
- Periodic human interventions may be necessary
- Set maximum iteration limits to prevent infinite loops
Combined Workflows
Real-world workflows often combine multiple patterns to achieve complex goals. Here's an example of a spec-driven development pipeline:
Spec-Driven Development Pipeline
Spec → (Modules) [Hierarchical]
↓
Each Module → Code [0→1 Creation]
↓
Code → Tests [1→1 Transformation]
↓
while(tests fail) → Fix [Error Correction]
Quick Reference
| Pattern | Notation | Use Case | Cognotik Task |
|---|---|---|---|
| Content Creation | 0 → 1 |
Generate new files from descriptions | FileTask |
| Transformation | 1 → 1 |
Edit, refactor, translate | EditTask |
| Synthesis | (A+B+...) → C |
Combine multiple inputs | InquiryTask |
| Analogical | (A−B+C) → D |
Feature transfer | EditTask |
| Incremental | A → A+δ |
Continue/extend content | FileTask |
| Hierarchical | A → (B,C); ... |
Multi-stage generation | TaskPlanningTask |
| Iterative | for(1..n) A → A |
Repeated refinement | EditTask |
| Error Correction | while(fail) fix |
Validation-driven fixes | RunShellCommandTask |
Ready to Apply These Patterns?
Explore how Cognotik implements these use case patterns through its task planning system.