Making Kedro projects easier to work on using GenAI

[rashidakanchwala]

Based on our recent MCP Server investigation and RAG knowledge retrieval spike, here are some potential future explorations:

Future use cases

• Address common pain points in working with data.
• Auto-generate a Data Catalog from data inspection, including dataset type suggestions.
• Agentic Data Engineering — leverage Kedro’s rich metadata to improve pipelines and workflows.

Other opportunities

• Pipeline Builder + LLM for low-code/no-code Kedro project creation.
• Expand CLI commands — LLMs handle CLI-style prompts very well.
• Stronger testing/validation for GenAI-built Kedro projects (e.g., KedroStructuredSchema).