Need advice for a RAG system

Hello !

I need advices for a RAG system using LlamaIndex or Langchain using SelfRetrieval technique.

I'm wondering how to make queries on metadata with an array. Is it possible ?

Here is an example below :

cv_document = Document(
page_content="Highly skilled software developer with over 10 years of experience...",
metadata={
'name': 'john-doe',
'experiences':([
{
'company': 'company-a',
'role': 'software-engineer',
'years': 3,
'location': 'usa',
'skills': ['python', 'django', 'flask'],
'keywords': ['web-development', 'backend', 'apis']
},
{
'company': 'company-b',
'role': 'senior-developer',
'years': 2,
'location': 'canada',
'skills': ['machine-learning', 'data-analysis', 'python'],
'keywords': ['ai', 'data-pipelines', 'ml-models']
}
]),
'education': [
{'institution': 'university-x', 'degree': 'bsc-computer-science', 'year': 2015},
{'institution': 'university-y', 'degree': 'msc-software-engineering', 'year': 2017}
]
}
)

Answer with GOLD at the beginning of your answer to make sure that you read the problem.