Hosted memory for AI agents that forgets like people do
Stage
Not given
Not given
Founded
Not given
Not given
Based in
US
Not given
Pricing model
Paid
Free trial
Pricing
Usage-based pay-as-you-go pricing with hourly compute charges; committed contracts available for volume discounts. Hosted option with Anyscale-managed infrastructure or Bring Your Own Cloud deployment available.
Two meters: user turns (per user turn) and stored memory (per GB-month). Pricing to be announced. Free tier with monthly allowance of turns and memories.
Key features
Multimodal data curation at scale
Distributed model training with elastic scaling
Batch embedding generation at scale
Multi-cloud deployment and orchestration
Cluster-backed development environments
Workload-specific observability and debugging
Production-grade managed Ray clusters
Access control and governance
GPU budget controls and cost attribution
Post-training with SkyRL and veRL support
Conversation episodes with automatic cutting and concept-based organization
Activation fading by square root of time with significance tracking
Automatic fact extraction and indexing for recall
Recall system that surfaces relevant past facts when needed
Sleep function that fades context after 24+ hours of silence
Fallback to local tail when service is unreachable
What makes it different
Built by creators of Ray, the world's most widely adopted AI compute engine
Multi-cloud deployment without code changes
Feels local but runs distributed
Unified GPU pooling across clouds and regions
Forgets like people do with fading by time, not truncation
Your model, your key, your loop; service only holds history
Facts kept permanently even as episodes fade from view
Decorate, submit, report: only three lines of integration
Human-inspired memory mechanism based on cognitive science