Binary and Scalar Embedding Quantization for Significantly Faster & Cheaper Retrieval - Tech Sentiments