Spell Teams Plan | $329 per user per month
This plan is for AI teams, including AI startups, corporate R&D labs, and other collaborative teams developing and deploying deep learning applications on the customer’s AWS, GCP, or Azure cloud.
Use all available GPUs for model training and serving, with minimal DevOps support
Enable secure, auditable, transparent deep learning workflows for production models
Build and train large models requiring distributed training and complex orchestration
Track, manage, document, and report all aspects of collaborative experimentation
Transparent model version roll-back for automated retraining and optimization
Configure and serve production models for robust, scalable deployment
Capture complete model lineage for reproducibility, governance, and accounting
Spell integrates and deploys infrastructural setup within an existing cloud account, creating a brand new VPC to contain all compute usage, ensuring maximum security and integration.
Multi-user experiment management, including project organization, experiments comparison, metrics tracking, and more.
Serve trained models on endpoints back by a managed Kubernetes cluster, and integrated with Grafana monitoring.
Continuously-monitored hotline to Spell engineering for project help, guidance, and review
Access to numerous in-depth articles and tutorials on AI infrastructure, model performance management and analysis, and more
The Spell Developer plan provides AI teams with everything they need to develop, train, deploy, manage, and track deep learning models in a production environment for AI startups and enterprises