feat(generator): add pipeline stage performance metrics

[kunalbhardwaj2006]

Summary

This PR introduces a lightweight performance metrics utility for the generator pipeline.

The goal is to measure execution time for different pipeline stages so developers can better understand the performance characteristics of the generator.

---

Implementation

A new utility class has been added:

generator/src/metrics.py

The PipelineMetrics class allows developers to track the execution time of different pipeline stages using:

metrics.start(stage_name)
metrics.end(stage_name)

At the end of execution, a report can be generated to show how long each stage took.

Example output:

PDF Download completed in 5.2 seconds
Question Extraction completed in 18.7 seconds
Classification completed in 92.3 seconds

---

Benefits

• Improves observability of the generator pipeline
• Helps identify performance bottlenecks
• Makes future optimization easier

---

Future Improvements

This metrics system could be extended with:

• structured logging
• exporting metrics to monitoring tools
• visualizing pipeline execution performance

---

Feedback and suggestions are welcome.

[kunalbhardwaj2006]

Hi @bruno 👋

This PR introduces a small metrics utility to track execution time of generator pipeline stages.

The goal is to make it easier to identify slow steps in the asset generation process and help optimize the pipeline.

I'd appreciate any feedback or suggestions for improving this approach.