Vector Databases

In this episode, we will explore Vector databases, a cutting-edge technology revolutionizing data storage and retrieval. We'll examine how these systems efficiently handle high-dimensional data, enabling advanced search capabilities and powering modern AI applications across various industries.

Guests

• Oumayma Essarhi
• Taha Bouhsine
• Merouane Zouaid

Notes

0:00:00 - Introduction and welcoming

0:03:47 - What is a Vector database?

0:13:47 - Why using a Vector Database, and what are the differences with classic Databases?

0:19:15 - Techniques used to speed search with Vector Databases

0:31:58 - Vector databases and AI

0:46:30 - Dealing with "curse of dimensionality" in Vector Databases

0:55:02 - Privacy concerns while using Vector databases

1:03:56 - Scalability of Vector Databases

1:12:54 - Q&A and Giveaay

1:27:00 - Conclusion

Links

• Extracting Training Data from Large Language Models
• 30 days of ML
• Strategy of pickinga Vector database
• PGrounding AI in reality with a little help from Data Commons
• DataGemma
• Masked Image Modeling with Vector-Quantized Visual Tokenizers
• An Any-to-Any Vision Model for Tens of Tasks and Modalities

Prepared and Presented by

• Meriem Zaid