Auto Evals annotations
Prompt engineering to get predictable code solutions from an llm ( full applications ). Creating an automated flow to extract the code run it and run end to end evals on it to analyze model accuracy. (multiple frameworks javascript, python, fullstack)