EloqData provides Redis‑compatible (EloqKV) and MongoDB‑compatible (EloqDoc) data stores that keep hot data on NVMe SSDs and use object storage for durability, delivering DRAM‑level throughput and P99.99 latency at a fraction of traditional DRAM‑based costs. Its Data Substrate architecture enables constant‑time branching and rollbacks on terabyte‑scale datasets, allowing developers to scale storage independently from compute while maintaining consistency and high availability.
Funding
Funding not disclosed
Founders
Product
Problem
Enterprises face escalating costs and scalability limits when using DRAM‑based caches and traditional cloud databases, which also suffer from high tail latency on SSDs and expensive block‑storage architectures. This hampers performance for AI‑driven workloads that require massive, low‑latency data access.
Solution
EloqData offers a suite of Redis‑compatible (EloqKV) and MongoDB‑compatible (EloqDoc) data stores built on its Data Substrate architecture. By decoupling compute, memory, logging, and storage and placing hot data on NVMe SSDs while using object storage for durability, the platform delivers DRAM‑level throughput and P99.99 latency at a fraction of the cost. The system provides durable, consistent, and globally addressable state for agents and AI applications, supporting fast, constant‑time branching of terabyte‑scale datasets. EloqCloud delivers these capabilities as a managed service with production‑grade free tiers, enabling developers to scale storage independently from compute while maintaining high performance and compliance.
Target Audience
Primary customers are cloud‑native developers, AI platform engineers, and enterprises that require high‑performance caching or document databases at scale, especially those building agent‑based or real‑time AI applications.
Features
- Redis‑compatible key‑value store (EloqKV) that keeps hot data on NVMe, achieving DRAM‑level throughput and sub‑millisecond P99.99 tail latency
- MongoDB‑compatible document store (EloqDoc) with tiered storage (memory, NVMe, object) delivering up to 10× higher throughput than comparable free tiers
- Data Substrate architecture that fully decouples compute, memory, logging, and storage, allowing independent scaling of each resource
- Constant‑time branching and rollbacks on TB‑scale datasets, enabling “Git‑like” experimentation on production data
- Truly distributed design that scales from a single node to thousands without sacrificing consistency or efficiency
- SOC 2‑audited EloqCloud service with built‑in high availability, cross‑AZ object‑storage durability, and automatic scale‑to‑zero