FAQs
Gathr Unlimited (On-prem/ VPC)
How can I register for a trial?
Do you extend support during the trial?
Do I need to use a professional organisation ID to sign-up?
Can I sign-up using a social service ID?
What sources or connectors does Gathr support?
How can I connect to sources other than the ones available out of the box?
Is it possible to ‘route’ messages to different operators based on some criteria?
Can I integrate multiple streaming pipelines?
Do you support complex event processing like pattern detection or correlation of events from multiple sources?
I want to calculate statistical functions over a moving window in real-time. Can this product help me?
Can I execute my custom business logic in a pipeline?
What is the installation process?
If I already have a Hadoop cluster, can I connect to it?
If I already have a kafka cluster, can I connect to it?
How do I migrate/move from pipeline from Dev to Production Environment?
What all cloud vendors are supported by Gathr?
How do I reach out to the professional services team for Gathr?
Do you provide customizations in your software?
Do you provide support for use-case development?
When should I use Gathr?
Do you support multi-tenancy?
Does Gathr Support Complex Event Processing?
Gathr has built in support for Complex Event Processing. The CEP operators can be used for pattern matching using “SQL like” queries.
Developers can define temporal patterns such as A followed by B followed by C or detect absence of an event in a pattern etc. Developers can write queries either using the CustomCEP processor or register a query through REST UI to enable dynamic configuration of queries and actions such as update/delete.
The user can also trigger actions based on the output of such temporal queries.
What kind of Machine Learning Capability does Gathr have?
Gathr has robust set of tools to build and operationalise machine learning models at scale. Following are the advanced features supported by Gathr play a key role in building a robust analytics solutions
- Visual interface of data preparation and representation of data in a graphical pane to derive correlation and feature engineering
- Training Models on multiple prominent technologies like SparkML, Notebooks, Python, H2O and similar tools
- Model Testing by distributing data into various training and testing sets, and model performance interface with multiple indicators to give a holistic view of various performance metrics and in depth view of model's performance.
- Hyper parameter tuning to tweak important metrics of the model and fine tune the output accuracy
- Workflows in Gathr enable scheduling of training pipelines on desired frequency for continuously trained machine learning models
- Automated deployment of models with Single click deployment of models for scoring
- Trained models can be very easily dropped on the canvas for inference in scoring pipelines for anomaly detection
- Support for model management, versioning and retraining
- Drift Detection to track deteriorating model performance and alerting users for appropriate actions
- Model Snapshotting and Reproducibility for managing history and auditing