GraphLake

An open file format and graph database engine

The GraphLake database engine is a OneGraph (unified RDF and property graph), scalable, high-performance, highly parallelized graph engine. Both ingestion and query are quick. It utilizes the GraphLake file format to allow efficient storage, update, and processing of graph data. As it is built on files and MPP architecture, it allows the storage and compute to be scaled independently.

Engine Features

Support for OneGraph (RDF, Property Graph)

Unified Analytics and OLTP workload support

Embedded Javascript Query Engine

Semantic OpenCypher - draft

SPARQL query language - subset for now

Highly parallelized query

Highly parallelized ingestion - BSBM 100m in ~4mins

Smart, continuous, compaction with no hang

Point in Time queries

Support for multiple storage backends (S3, SSD, Azure)

Automated and Intelligent Sharding

Free for developers - Forever!

Competively priced for business

Read the docs

Open File Format for Graph Data

Inspired by Apache Iceberg we wanted a file format for native graph analytics and processing.

Learn more about the GraphLake open file format.

View the Spec

Versions & Pricing

Note. All offerings are to be run in your cloud or your hardware.

Features Developer Businees Business Plus Enterprise
 Price Free as Docker Image / Binary coming soon coming soon coming soon
 Connections 1 10 Unlimited Unlimited
 Memory Limit Unlimited Max 16Gb on host Max 32Gb on host Over 32Gb on host
 Storage Backends Local Filesystem / Mounts Local Filesystem / Mounts Local, S3, Azure Blob Local, S3, Azure Blob
 Compaction No Yes - limited schedule Yes - continuous Yes - continuous
 PiT Queries Yes No Yes Yes
 User Management No - single user unsecured Yes - 5 user accounts - secured requests Yes - 25 user accounts - secured requests Yes - Unlimited user accounts - secured requests
 Multi Graph Yes Yes Yes Yes
 Data Branching No No Yes Yes
 Support Community Email - best effort Email 48hr Response Email 16hr Response (other levels available)
Download Buy One Click Deploy

MultiNode

GraphLake scales up great on a single box, especially when paired with high performance object storage. However, we think that having multi-node on compute will take things even further. The MultiNode version of the engine is in development. If you would like to find out more, please contact us.