OpenAI Data Analysis Specialist

I need to push the OpenAI API far beyond simple prompt-response calls and understand exactly how to harness it for solid, repeatable data analysis. The goal is a clear, well-documented workflow (and companion prototype code) that shows:

• how to structure raw text data, stream it through the right OpenAI models and parameters,
• how to evaluate accuracy, latency, token costs and scalability,
• and how to package the whole pipeline so it can drop into a larger analytics stack written in Python.

You’ll run controlled experiments, capture metrics, comment the code thoroughly and summarise every finding in a concise technical report. By the end I want a working notebook or script I can extend, plus your recommendations for model selection, prompt engineering, rate-limit handling and future fine-tuning paths.