A developer's agent needs to process CSV files uploaded by users. The files vary from 100 rows to 1 million…

Question

A developer's agent needs to process CSV files uploaded by users. The files vary from 100 rows to 1 million rows. For small files, direct context inclusion works well. For large files, the context can't hold all data. What scalable architecture handles both cases?

Accepted Answer

C. For small files (<5K rows): include directly in context. For large files: load into a database and provide SQL query tools for the agent to explore the data on demand

Answer

A. Always process the full CSV regardless of size, relying on compaction if needed

Answer

B. Always chunk the CSV into 100-row segments and process sequentially

Answer

D. Convert all CSVs to summary statistics before processing

Question 276