The secret to your AI/ML success
Reaching your ROI is no longer a challenge when you use Qwak for all your AI Infra and MLOps needs. Pay only for what you use with our affordable plans.
Free
Best suited for dev environments
Start FreeUp to 100 QPU/month for a year
Unlimited model versions on Qwak model registry
Train and deploy on scalable resources
Managed Jupyter notebooks
Transform, store and manage features in the Feature store
Automatic inference and training pipelines
Monitor models and features
Model A/B testing and traffic splitting
Up to 1GB in-memory and 10GB storage
Chat support
Pay As You Go
Best suited for productions environments
Get StartedEverything in Free, plus:
Unlimited QPU
Multiple environments
<2 hours SLA for critical cases
99.8% uptime guarantee
Multi A-Z installation (DR)
Enterprise
Best suited for enterprise environments
Contact UsEverything in Pay As You Go, plus:
Volume discounts for pre-commitments
Self hosted hybrid deployments
SSO login
RBAC support
Dedicated support channel
<1 hour SLA for critical cases
<2 hours for important cases
Dedicated Qwak architect hours/month
2 annual on-site workshops
99.9% uptime guarantee
Private VPC installation
Storage
The Qwak platform contains two types of storage. Super fast in memory storage which is primarily used for online feature store, and warehouse storage, used for offline feature stores, model analytics, and monitoring.
Qwak GPU Cloud
Achieve top model performance and save costs with effortless GPU autoscaling and simple model deployment on any of the various GPUs available on the Qwak GPU cloud.
The world’s best AI teams use Qwak
FAQs
The platform is made for data scientists and AI practitioners who are looking to eliminate dependency on engineering resources and ML engineers who need someone else to take care of the heavy lifting so they can focus on their business’ needs.
We're a usage-based platform measured in QPUs, an internal processing unit we use at Qwak. We charge for build and deployment time, along with the resources you choose to use.
QPU, or Qwak Processing Unit, is our internal processing unit. To see how QPU’s are calculated, please see our documentation.
All Python based models are supported.
We support all common data sources, such as Snowflake, Kafka, BigQuery, RedShift, SQL, etc. We do not support data sources that are on-premise. See our documentation for the full list of supported data sources.
Qwak takes the confidentiality and integrity of its customer data very seriously and strives to ensure that data is protected from unauthorized access and is available when needed. Additionally, our assumption is that all data added into the platform by our customers is potentially PII / sensitive data, and is therefore treated as such.