Add advanced examples: pro-dev quick start and data science risk assessment

Pro-dev quick start (prodev_quick_start.py):
- Create 4 custom tables with columns and relationships
- Populate via DataFrame CRUD, query, analyze, update, delete, cleanup
- Uses result.primary_name_attribute for correct column naming
- Includes pandas vs dict/list rationale

Data science risk assessment (datascience_risk_assessment.py):
- 5-step pipeline: Extract -> Analyze -> LLM Summarize -> Write-back -> Report
- Statistical risk scoring with pandas/numpy (groupby, merge, percentile)
- 3 LLM provider options: Azure AI Inference, OpenAI, GitHub Copilot SDK
- Template-based fallback when no LLM configured
- LLM interaction logging (prompts, responses, timing)
- Optional matplotlib visualization with chart export
- CSV data exports to script-relative output folder

Author: Saurabh Badenkal