Databricks Inc.

Instabase, Inc.

Clemens Mewald

Aaron Daniel Davidson

Tomas Nykodym

Matei Zaharia

Shaunak Godbole

Shih Ping Chang

Anant Bhardwaj

Mani Parkhe

Avesh Singh

Systems and methods for providing access control for federated retrieval-augmented generation (FRAG) using a set of vector databases are disclosed. Exemplary implementations may: receive user input representing a user query; verify whether the user has access to particular electronic files; generate agent queries based on tasks included in the user query; select subsets of the vector embeddings based on the agent queries; provide the subsets to one or more machine learning models as context; provide prompts to the one or more machine learning models; obtain replies in response to the prompts; generate a reply to the user query; effectuate a presentation of the reply, and/or perform other steps.

Systems and methods for access control for federated retrieval-augmented generation

Systems and methods for controlling retrieval-augmented generation (RAG) using a set of vector databases are disclosed. Exemplary implementations may: convert content within sets of electronic files into sets of vector embeddings stored in vector databases; receive user input representing a user query; generate agent queries based on tasks included in the user query; select subsets of the vector embeddings based on the agent queries; provide the subsets to one or more machine learning models as context; provide prompts to the one or more machine learning models; obtain replies in response to the prompts; generate a reply to the user query; effectuate a presentation of the reply, and/or perform other steps.

Systems and methods for agent-controlled federated retrieval-augmented generation

The present application discloses a method, system, and computer system for managing a plurality of features and storing lineage information pertaining to the features. The method includes obtaining one or more datasets, determining a first feature, wherein the first feature is determined based at least in part on the one or more datasets, and storing the first feature in a feature store. The first feature is stored in association with a dataset indication of the one or more datasets from which the first feature is determined. The feature store comprises a plurality of features.

Feature store with integrated tracking

A system includes an interface, a processor, and a memory. The interface is configured to receive a version of a model from a model registry. The processor is configured to store the version of the model, start a process running the version of the model, and update a proxy with version information associated with the version of the model, wherein the updated proxy indicates to redirect an indication to invoke the version of the model to the process. The memory is coupled to the processor and configured to provide the processor with instructions.

Growing community of inventors

Lafayette, CA, United States of America

Clemens Mewald