Ram Sriharsha on a new vector store built with the Rust programming language.
Ram Sriharsha is VP of Engineering and R&D at Pinecone, a startup that offers a fully managed vector database (not just an index). While vector databases are used for recommendation, anomaly detection, and Q&A systems, they primarily target search and information retrieval applications. If your organization is using AI, chances are you have embeddings. In some sense, vector databases are what you get when you embed your entire database. (For more on Vector Databases, see our recent primer.) In this episode, we discuss Pinecone’s new proprietary storage engine, which was first described around the time we recorded this conversation.
Highlights in the video version:
- Introduction to Ram Sriharsha
Vector databases, vector embeddings, and case studies
Relationship between vector databases and graph databases and feature stores
Vector database storage capabilities and optimal way of indexing
Compressing data and scaling out strategies
Fault tolerance, enterprise features, and the new update
What do you provide around lineage and reproducibility?
What’s on the roadmap for vector databases?
Search engine query and industrial applications
- A video version of this conversation is available on our YouTube channel.
- The Vector Database Index
- Summer of Orchestration: conversations with co-creators of Prefect, Dagster, Flyte, and Orchest.
- New open source tools to unlock speech and audio data
- fastdup: Introducing a new free tool for curating image datasets at scale
- A Guide to Data Annotation and Synthetic Data Generation Tools
If you enjoyed this episode, please support our work by encouraging your friends and colleagues to subscribe to our newsletter:
[Image: Embeddings by Ben Lorica, using images from Infogram.]