Spiral provides a transactional multimodal database that natively stores embeddings, images, video and other data types while integrating directly with object storage. It delivers unified permission management and a hybrid deployment model that spans public clouds, private clouds and on‑prem clusters, enabling high‑throughput pipelines that align multi‑rate streams and saturate GPUs for large‑scale AI training and analytics.
Funding
$20.5M 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.
Founders
Product
Problem
Organizations building large‑scale AI models must handle massive multimodal datasets—embeddings, images, video, and other streams—while keeping data accessible, secure, and performant across cloud and on‑prem environments. Existing storage and database solutions often create bottlenecks, require separate systems for analytics and training, and lack unified permission management.
Solution
Spiral offers a single transactional, multimodal database that natively stores embeddings, images, videos, and other data types while being backed by object storage. The platform aligns multi‑rate data streams, enables dynamic construction of training sets, and delivers machine‑scale throughput to fully saturate GPUs. Permissions are unified across all data, and the system runs in a hybrid mode that spans hyperscalers, private clouds, and on‑prem clusters, eliminating data silos and simplifying deployment for both analytics and model training workloads.
Target Audience
Primary customers are frontier AI research labs, large‑scale machine‑learning teams, and enterprises that need a unified data platform for training, analytics, and inference at machine‑scale throughput.
Features
- Transactional multimodal database supporting embeddings, images, video, and arbitrary data types
- Direct integration with object storage, providing scalable capacity and low‑cost durability
- Unified permission model applied consistently across all stored data
- Hybrid deployment architecture that operates across public clouds, private clouds, and on‑prem clusters
- High‑throughput data pipelines that align multi‑rate streams for dynamic training‑set generation
- Optimized for GPU saturation, delivering machine‑scale data delivery for large AI workloads