ScyllaDB is a distributed NoSQL database that uses a shard‑per‑core architecture and lock‑free communication to deliver single‑digit millisecond latencies at multi‑million‑operation workloads. It provides active‑active replication, automatic tuning, and a unified in‑memory cache, and can be deployed on‑premises or accessed as a managed cloud service compatible with Cassandra CQL.
Funding
$43M raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
ERFounders
Product
Problem
Enterprises that require real‑time data processing often encounter latency spikes, unpredictable performance, and high operational costs with traditional NoSQL solutions such as Cassandra, DynamoDB, or MongoDB, especially when workloads scale to millions of operations per second. These limitations make it difficult to meet strict service‑level agreements and to control total cost of ownership at scale.
Solution
ScyllaDB delivers a distributed NoSQL database built on a close‑to‑the‑metal architecture that extracts the full potential of modern multi‑core servers. By assigning a dedicated shard to each CPU core and employing an asynchronous, lock‑free communication model, the system sustains single‑digit millisecond P99 latencies even under multi‑million‑ops workloads. Integrated components such as a unified cache, latency‑aware I/O scheduler, and self‑tuning configuration minimize administrative overhead while guaranteeing predictable performance. The platform supports active‑active replication for high availability and can be deployed on‑premises or accessed as a fully managed cloud service, allowing customers to scale horizontally with linear cost and operational efficiency.
Target Audience
The primary customers are large‑scale application developers and engineering teams in sectors such as e‑commerce, ad tech, IoT, gaming, and telecommunications that need high‑throughput, low‑latency NoSQL storage, as well as cloud providers and enterprises seeking a managed DBaaS solution.
Features
- Shard‑per‑core design that isolates each core as an independent resource, eliminating lock contention and maximizing CPU utilization.
- Unified in‑memory cache that provides ultra‑low latency reads without requiring external caching layers.
- Adaptive I/O scheduler that prioritizes latency‑sensitive tasks and auto‑tunes concurrency for peak throughput.
- Active‑active replication with real‑time data synchronization to ensure continuous availability and fault tolerance.
- Linear horizontal scalability across nodes, enabling seamless scale‑out for global workloads.
- Low‑touch autotuning that configures optimal settings at startup, reducing manual tuning effort.
- ScyllaDB Cloud (DBaaS) offering with usage‑based pricing and built‑in monitoring, simplifying operations for teams that prefer a managed service.
- Compatibility with Cassandra query language (CQL) and native drivers, facilitating migration from existing NoSQL stacks.