Debug Google Vision Python System

I need help debugging a Google Vision recognition system built with Python. A previous develper created a solution for me ovr 4 years ago and it has stopped working. I do not know the reason and can't contact the original develper. I believe the models are still created correctly but the code to return the 'matches' is no longer working.

Ideal skills and experience:
- Proficiency in Python
- Experience with Google Vision API
- Strong debugging skills
- Familiarity with image processing and analysis

Please provide a brief overview of your experience and approach to this issue.