Find all the AI needs in one place.
A guide to using delimiters in prompt engineering
""
Enclosing specific text segments
Translate this text: "Hello world"
[], {}, ()
Lists, options, or explanations
Ingredients: [flour, sugar, eggs]
|, #, @
Separating different parts
Step 1 | Step 2 | Step 3
New lines, tabs
Visual separation
Instructions: 1. First step 2. Second step
Use consistent delimiters throughout your prompt
Choose simple, recognizable delimiters
Select context-appropriate delimiters
Test and iterate based on results
Avoid nested delimiters when possible
❌ Without Delimiters:
Generate a summary about AI
✅ With Delimiters:
Generate a summary about: "Artificial Intelligence"
❌ Without Delimiters:
List ingredients flour sugar eggs
✅ With Delimiters:
List ingredients: [flour, sugar, eggs]
❌ Without Delimiters:
Translate to French then summarize The cat sat on the mat
✅ With Delimiters:
Translate to French: "The cat sat on the mat" Then summarize the translation.
A systematic approach to gathering, validating, and optimizing data for various types of generative AI models. This methodology ensures high-quality inputs for better AI outputs.
Use Case: Documentation, reports, analysis
Data Prep: Convert to clean text chunks, remove formatting
Use Case: Combined text, image, and data analysis
Data Prep: Align different data types, maintain relationships
Use Case: Speech recognition, audio analysis
Data Prep: Clean audio transcripts, timestamp alignment
Use Case: Video understanding, scene analysis
Data Prep: Frame extraction, scene descriptions
Use Case: Visual content creation, image editing
Data Prep: Image descriptions, style references
Determine the most appropriate type of generative AI based on your project needs
Assess project requirements and desired outputs
Evaluate available AI model capabilities
Consider computational resources and limitations
Review model-specific data format requirements
Gather relevant data from multiple sources while maintaining quality
Identify authoritative data sources
Scrape data using appropriate tools
Collect documentation and reference materials
Organize raw data by type and relevance
Verify data accuracy using AI-powered citation checking
Use internet-connected AI for citation verification
Cross-reference multiple sources
Check data freshness and relevance
Document validation results
Transform data into the most effective format for the chosen AI
Format data according to model requirements
Chunk content appropriately
Add necessary context and metadata
Optimize for model performance
Focus on high-quality, validated data rather than large volumes of uncertain information. Well-prepared data leads to better AI outputs.
Regularly verify data accuracy using AI-powered citation checking and cross-referencing to maintain data quality.