Hey all we are using llama-index version 0.9.5 for an older project and we were wondering if there's a functionality in this version to query two different vectorstoreindices, and ensure that nodes from both are used to synthesize an answer. I can't find the queryfusionretriever in this verison of llama-index, and the closest I can find is building a custom retriever. Any ideas / approaches to tackle this issue?
