Play 47
Synthetic Data Factory
Privacy-safe synthetic data generation preserving statistical properties with zero real PII.
Privacy-safe synthetic dataset generation for AI training and testing — realistic tabular, text, and structured data preserving statistical properties with zero real PII, differential privacy validation, and bias detection. Uses Azure Machine Learning for generation pipelines and Blob Storage for versioned dataset management. Supports GDPR and CCPA compliance workflows.
Architecture Pattern
LLM-powered data generation: differential privacy validation, statistical fidelity scoring
Azure Services
DevKit (.github Agentic OS)
- agent.md — root orchestrator with builder→reviewer→tuner handoffs
- 3 agents — Synthetic Data Builder (gpt-4o), Reviewer (gpt-4o-mini), Tuner (gpt-4o-mini)
- 3 skills — deploy (246 lines), evaluate (146 lines), tune (227 lines)
- 4 prompts — /deploy, /test, /review, /evaluate with agent routing
- .vscode/mcp.json — FrootAI MCP with OpenAI + Storage inputs + envFile
TuneKit (AI Config)
- config/openai.json — gpt-4o for data synthesis
- config/generation.json — schema definitions, row count, distribution rules
- config/guardrails.json — differential privacy ε, PII detection
- evaluation/eval.py — Statistical fidelity >90%, PII leak rate 0%
Tuning Parameters
Machine evidence
FrootAI evidence lifecycle
This is an internal evidence maturity label, not third-party certification, accreditation, legal compliance, or a production guarantee. Missing or expired evidence demotes automatically; catalog claims cannot promote a play.
This play currently has design evidence only. A runnable scenario, endpoint evaluation, and build receipts are the next contiguous gates.
Repo Intelligence
v1A no-clone, revision-pinned map for agents and humans. Observed evidence is separated from inferred flow so the output stays useful without pretending to be a full call graph.