Use the power of AI and ML to empower people, make data-driven decisions and take action for improved, strategic business outcomes. Torry Harris’ award-winning AI and ML-based analytics and prediction framework, 4Sight, is a user interface-based tool that creates predictive models without involving custom code and can be configured to receive forecasted values based on a set of input parameters. It is built to work with any data source (logs, databases, streaming data, and more). The framework generates widgets that can be embedded into any existing applications, seamlessly bringing analytics and prediction capabilities to any application.

With 4Sight, you can:

  • Map data from multiple sources and define custom fields
  • Create widgets that can be specific to users/teams
  • Use a UI-based approach to create prediction models
  • Integrate with applications via API or widgets
  • Support for different visualizations like line/pie/bar charts
Login
The login page is an entry page to 4Sight that requires a username/email and password to authenticate and proceed to the 4Sight's internal pages.
4Sight Landing Page
The 4Sight landing page displays all features and functions of the tool, intuitively designed to maximize user experience.
The All Widgets Page
Widgets allow users to generate a report dynamically. The “All Widgets Page” provides information about the type of chart along with the name. Clicking a widget item will open a dashboard that displays dimension, metrics, time range, and sorting.
The Dashboard
The Dashboard provides an overview of multiple widgets that have different configurations, simultaneously, on one page. It also provides users the ability to Search, Sort, Create a New Dashboard, View more, or Delete a Dashboard.
Search
The user can search for specific items on the dashboard by entering a key value or a phrase in the search bar. On the ALL SEARCH page, users can Search, Sort, Create a New Search, Export and Import Search Configure, View, More (Delete).
Prediction
This feature allows users to use available data to create and validate a model that can be used to make future predictions.
Dataset
The dataset module allows users to select the dataset for widget creation on the widget page. Here users have the flexibility to retain the same column name, or they can mask with a different name.
Widget canvas
On the WIDGET CANVAS page, users can restrict the exposure of the column in the widget and can mask the name. Whatever columns the user selects will be available in Widgets, Dashboard, Search, Prediction, etc.

Case Study

4Sight detects service anomalies invisible to the human eye

Read Now
  • Built to complement existing applications
  • Plug-in into different data sources
  • Define custom schemas
  • Supports multi-tenancy
  • Built in prediction models
  • Reduces duplication of data
  • Change prediction & retrain the model using the UI
  • Control the Analytics/Predictions from the UI
  • Define fine grained access to the data
  • Hide sensitive fields from logs
  • Inbuilt log search capability
  • Re-use existing Elastic Search setup
Why
4Sight?
4Sight AI / ML Capabilities
  • Data Forecast based on historical patterns
  • Recommendation system based on similar patterns
  • Predict & alert future failures/events
4Sight Log Search and Alert Capabilities
  • Supports time-based log search
  • Define multiple search tags based on user events
  • Alerts based on data patterns and time ranges
  • Configurable alert intervals
4Sight Prediction Models
  • UI driven prediction model configuration
  • Provides accuracy of trained models
  • Scheduled model training
4Sight Widget Creation
  • Drag and drop interface
  • Supports different visualization like pie, bar, line, area, donut, gauge.
  • Supports filters at different levels
  • Customizable time range
  • Supports custom sort
  • Supports custom script

It’s time to give your organization real foresight

Contact us

Resources and best practices

Blog
→
Bringing foresight to API management to leverage digital marketplaces
Blog
→
Implementing AIOps to predict runtime failures
Case Study
→
Not too small to fail: AIOps detect service anomalies invisible to the human eye