Pinecone emerges from stealth with $10 million to accelerate AI workloads

Pinecone Systems, a company developing a database platform for AI workloads, today emerged from stealth with $10 million in funding. A company spokesperson told VentureBeat the funds will be put toward customer acquisition and product R&D as Pinecone’s platform launches for self-onboarding.

AI models take data, whether images, sentences, or user behavior, and convert them into complex collections of numbers called vectors. Using a machine learning model is often a question of finding out which vectors are nearest or most similar to others, but existing ways of doing this can be slow and inaccurate.

Pinecone aims to make AI queries faster, more accurate, cheaper, and easier to use with fully managed serverless vector database technology. Founded by developers from Facebook and Google — along with CEO Edo Liberty, who ran Yahoo’s Scalable Machine Learning Platforms group and later Amazon AI Labs — Pinecone’s database supports deployments of personalization, semantic text search, image retrieval, data fusion, deduplication, recommendation, anomaly detection, and other real-time applications.

Developers can use Pinecone to configure and launch services with custom machine learning models, transformations, and rankers. They’re also able to add or update data in batches or streams; data is turned into vectors and the index is updated in real time. Moreover, they can monitor and track deployments for operational health and support queries to find similar or top-ranking items.

Liberty claims Pinecone, which connects to third-party data sources, can scale to billions of high-dimensional vectors while keeping latencies below 50 milliseconds for queries, updates, and embeddings with horizontal container distribution. Pinecone ostensibly has a real-time indexing speed 30 times higher than open source libraries.

“One of the largest retailers in the world reports using Pinecone to serve real-time shopping recommendations based on their own deep learning models. They saw an immediate 18.5% lift in revenue per recommendation compared with their previous solution,” a Pinecone spokesperson told VentureBeat. “Pinecone’s database is 100% serverless and API-driven, which means customers always have the computing resources they need, when they need them, without having to worry about infrastructure or maintenance.”

Wing Venture Capital led the seed investment in Sunnyvale, California-based Pinecone announced today.

“The modern enterprise is built on data and powered by AI,” Peter Wagner, a founding partner at Wing Venture Capital and the newest member of Pinecone’s board of directors, said in a statement. “The data cloud has emerged as its foundation with the ascendance of Snowflake. Pinecone is poised to unleash data teams and their ML-based applications in a similar fashion.”