Python CLI that normalizes a messy dataset of OptinMonster opt-in campaigns, benchmarks each site against its industry peers, and uses GPT to generate one validated, grounded next-best-action recommendation per site - deterministic code owns every number and decision; the LLM only reshapes prose.
python cli benchmarking openai recommendation-engine data-cleaning data-pipeline structured-output pydantic llm ninja-prototype ninja-interview-assignment
-
Updated
Jul 21, 2026 - Python