I am using pulumi to set up a function in lambda.

hello_world_lambda = aws.lambda_.Function(
    'hello-world-lambda',
    role=lambda_role.arn,
    runtime='python3.9',
    handler='lambda_function.handler', 
    code=pulumi.AssetArchive({
        '.': pulumi.FileArchive(f'./lambda/test/'),
    }),
    timeout=60
)

This succesfully uploads the content of my test dir to lambda. My function in lambda needs to utilize pandas (which is not native to lambda). If I install pandas directly in my test dir, it will fail when I try to import it in my lambda function. If instead I set a venv within my test dir, it will not be uploaded for whatever reason. What is the proper way to set up extra dependencies (in this case pandas) with pulumi in lambda?