Tl;dr - Summary of the points. Graph databases are here for long time Adaption rates are low. Graph databases are used as a relational database without tables. Getting data into graph databases is still a problem. Using Ontology to define relationships may work in scientific research not Enterprises. Ontology classifications can’t account for literally infinite amounts of variations and categories. Extracting relationships from LLMs is fraught with risks. If LLMs can ext

How can you make agents understand exactly and precisely what you say and do it? This is the burning question for everyone working with agents. Why do agents, and by extension LLMs misunderstand the prompt? Isn't AI supposed to understand clearly? Nope. That is fundamental problem in Linguistics not of AI. It comes down to the meaning of the word. When people read a word, how meaning is understood? By looking at the words near by. There is hypothesis in the Linguistics

In Information retrieval, similarity, the word invokes various meanings in various contexts. It is used in tasks where exact matching is not enough. In business cases where searching for information takes precedence, similarity search is the way to understand the trends, extract insights. In other cases, such as facial recognition, image comparison, etc similarity is a basic functionality requirement. Most of the business data sits as textual information. Most commonly consu

OpenSource search frameworks such as OpenSearch, Facebook AI Similarity Search (FAISS), Weaviate and Qdrant etc using vectors for similarity search. An improvement added via LLMs to create embeddings. Here vector means one dimensional numerical representation of data, embedding means multi-dimensional numerical representation of data. Though both are now interchangeably used, vectors are pre-LLM era usage. Let us see why vectors, embeddings are needed and why it still fails t