Hi all, Quick context: After exploring vector databases, I've pivoted to a more precise solution. Working with a tech expert, we've developed an agent using hardcoded Python script with SQLite that delivers 99.9% accurate results. Now, I need guidance on optimizing my Make.com workflow for my VAPI AI assistant setup. Current Flow: 1. VAPI triggers webhook in Make.com 2. Question transforms to JSON 3. Need to run Python code here (to query SQLite database) 4. Convert Python results to JSON 5. Send webhook response back to VAPI Specific Challenge: • Looking for the most efficient Make.com node configuration to: • Execute Python code • Handle SQLite database query (that is something that is already setup in the background) • Format results properly I've got the webhook trigger and JSON transformation working smoothly - just need help with the Python execution piece. *Has anyone implemented something similar? * Would appreciate any best practices or node recommendations!