Hyperscience R37 Release
Onboard Users More Easily
Flatten the learning curve for AI adoption, helping business users easily train data models.
Deploy New Business Cases
Support additional use cases that require more nuanced intent, topic, or sentiment classification.
Maximize Value
Deploy AI-powered automation at a significant scale, in a way that adapts to changes.
Fast, Efficient Enterprise Model Training
Reduce model training time by 37% on average, increasing user adoption and accelerating time-to-value.
- Train highly accurate models with less data, and receive guidance on which data provides the best ground truth
- Improve model accuracy by easily finding and fixing annotation errors
- Gain greater transparency into each model with details about which data was used for training
Fast, Efficient Enterprise Model Training
Reduce model training time by 37% on average, increasing user adoption and accelerating time-to-value.
- Train highly accurate models with less data, and receive guidance on which data provides the best ground truth
- Improve model accuracy by easily finding and fixing annotation errors
- Gain greater transparency into each model with details about which data was used for training
Effortless Enterprise Scale
Scale the use of AI to meet the changing demands of your organization without worrying about limitations in processing power, even during workload spikes.
- Reduce time-to-value with automatic system scaling to support process spikes—no more missed SLAs or document volume projections*
- Create SLAs that work in your business context by specifying deadlines for submissions received on business days and non-business days
- Expand the use of AI across your organization by enabling faster training for unstructured use cases
* Currently available for SaaS only, availability for on-prem deployments coming soon.
Learn more
Effortless Enterprise Scale
Scale the use of AI to meet the changing demands of your organization without worrying about limitations in processing power, even during workload spikes.
- Reduce time-to-value with automatic system scaling to support process spikes—no more missed SLAs or document volume projections*
- Create SLAs that work in your business context by specifying deadlines for submissions received on business days and non-business days
- Expand the use of AI across your organization by enabling faster training for unstructured use cases
* Currently available for SaaS only, availability for on-prem deployments coming soon.
Learn moreImproved Supervision and Routing Experience
Improve employee productivity by optimizing submission routing, and empower knowledge workers to make better and faster decisions.
- Improve the employee experience with a more flexible, richer supervision interface
- Achieve end-to-end document processing automation with enhanced routing tasks—providing keyers with more information for making decisions
- Expand adoption to use cases where multi-level classification by intent, topic, or sentiment is needed (E.g. email triage)
Improved Supervision and Routing Experience
Improve employee productivity by optimizing submission routing, and empower knowledge workers to make better and faster decisions.
- Improve the employee experience with a more flexible, richer supervision interface
- Achieve end-to-end document processing automation with enhanced routing tasks—providing keyers with more information for making decisions
- Expand adoption to use cases where multi-level classification by intent, topic, or sentiment is needed (E.g. email triage)
| Theme | Improvements |
|---|---|
| Simplify training data management | Improved filtering for identifying ineligible training data Easily identify and fix problematic training data that is ineligible to be included in the model training process |
| Table ID Anomaly Detection Anomaly Detection support has been extended to table models, which enables easy resolution of inconsistencies in annotation |
|
| Model Validation Tasks in Training Data Management Address disagreements between model predictions and annotation data from the Model page. |
|
| Effortless enterprise scale | GPU Support for On Prem Train more powerful unstructured models with Hyperscience deployed directly on your infrastructure. |
| Auto-scaling for SaaS Our infrastructure scales with your business needs whether you are processing 1 or 1,000,000 documents |
|
| SLA Management with Working Days Define and create processing SLA policies around the business days for your organization |
|
| Richer, easier user experience | Improved Flows Management More intuitively find and manage flows in the Hyperscience platform |
| (Private Beta) Out of the Box Models Get immediate value from Hyperscience by using out of the box models for specific document types Please reach out to your Hyperscience rep if you want to learn more! |
|
| Read the full details in the Release Notes | |