Agile application development

1. Zero data loss writes and low-latency, always-on reads

Default single-row ACID operations with optional distributed ACID transactions when needed. Multiple read consistency levels to serve latency-sensitive applications.

3. High performance

A ground-up C++ implementation combined with 3x read throughput compared to simple quorum-based consistency of traditional NoSQL.

Choose between 2 popular NoSQL APIs rather than learn a new proprietary language, even port existing applications fast by leveraging the maturity of these ecosystems. SQL support is in the works.

4. Converged caching

Avoid explicit management of the memory and consistency associated with an independent cache that typically fronts a persistent database.

Cloud-native technical operations

1. Multi-region/multi-cloud deployments simplified

Geo-redundant deployments across multiple availability zones and regions on public/private/hybrid clouds as well as on-premises datacenters.

2. Zero downtime and zero data loss operations

Highly available under any unplanned infrastructure failure as well as any planned software or hardware upgrade.

3. Simple scalability

Linear, fast & reliable scalability from 1 to 1000s nodes with automatic sharding and rebalancing. Use Kubernetes for orchestration of dev/test workloads.

4. Hardware flexibility

Seamlessly move from one type of compute and storage to another for cost and performance reasons.