User-centric and LLM-enhanced adaptive ETL code synthesis

A non-transitory computer readable medium, storing instructions for program code generation for data processing programs, the instructions comprising:

• for receipt of a user prompt:
• referencing a prompt database that associates historical prompts, program code generated from the historical prompts, and information indicative of whether execution results of the program code were expected or not expected to generate a combined prompt from the user prompt and a historical prompt from the historical prompts determined to be related to the combined prompt from the referencing;
• using historical prompts associated with program code that did not produce expected results to avoid generating similar unsuccessful program code;
• referencing a program template database that associates combinations of the historical prompts and historical program code that was verified to work as expected to generate program code with the combined prompt;
• and
• executing the generated program code and the combined prompt for the data processing programs,
• wherein the data processing programs are part of an extract, transfer, load (ETL) system, and
• wherein the prompt database stores user evaluations indicating whether generated ETL program code produced expected or unexpected results, and
• wherein the program code is iteratively regenerated using historical prompts associated with unexpected results until expected results are achieved.

An apparatus, storing instructions for program code generation for data processing programs, comprising:

• a processor, configured to, for receipt of a user prompt:
• reference a prompt database that associates historical prompts, program code generated from the historical prompts, and information indicative of whether execution results of the program code were expected or not expected to generate a combined prompt from the user prompt and a historical prompt from the historical prompts determined to be related to the combined prompt from the referencing of the prompt database;
• use historical prompts associated with program code that did not produce expected results to avoid generating similar unsuccessful program code;
• reference a program template database that associates combinations of the historical prompts and historical program code that was verified to work as expected to generate program code with the combined prompt;
• and
• execute the generated program code and the combined prompt for the data processing programs,
• wherein the data processing programs are part of an extract, transfer, load (ETL) system,
• wherein the prompt database stores user evaluations indicating whether generated ETL program code produced expected or unexpected results, and
• wherein the program code is iteratively regenerated using historical prompts associated with unexpected results until expected results are achieved.