Prompt file imported from ip2cloud/bravo-edd (.codex/prompts/ai-pipeline.md). Copyright stays with the author.

AI Pipeline Command

Scaffold RAG pipelines, embedding workflows, feature stores, and text-to-SQL

Usage

/ai-pipeline <description-or-file>

Examples

/ai-pipeline "RAG pipeline for internal docs with pgvector"
/ai-pipeline "Embedding pipeline from S3 PDFs to Pinecone"
/ai-pipeline "Feature store setup with Feast for ML models"
/ai-pipeline "Text-to-SQL agent for analytics queries"

What This Command Does

1. Invokes the ai-data-engineer agent
2. Analyzes your AI/ML data requirements
3. Loads KB patterns from ai-data-engineering and streaming domains
4. Generates:
   • RAG pipeline architecture and code
   • Embedding pipeline with chunking strategies
   • Vector database setup and indexing
   • Feature store definitions
   • Text-to-SQL prompt templates

Agent Delegation

KB Domains Used

• ai-data-engineering — RAG pipelines, vector databases, feature stores, LLMOps
• streaming — real-time embedding ingestion
• data-quality — embedding quality, drift detection

Output

The agent generates pipeline code, configuration, and architecture documentation for your AI data workflow.