These are the main features that we will enable soon:
- Anomaly detection
- Sub-Dimensional Anomaly Alert
- Enable Anomaly w/o Sub-Dimension
- Code/API level configuration for
- Hyperparameters for any algorithms
- Subpopulation calculation logic
- Alerting
- Backoff for alerting to minimize repeated alerts
We will be enabling the following new connectors soon:
- Redshift
- Hubspot
- Salesforce
- Google Analytics Custom Report (will likely already enabled)
- CSV
We are working on the following to improve & support data at larger scale:
- Data ingestion
- Support for materialized views across dataware houses for large scale processing
- Support for OLAP cube support where available
- Data processing
- Distributed Sub-Dimensional Anomaly Detection
- Higher level of sub population support
- Robustness
- More Robust & Fault Tolerant Task Scheduler
- Robust Error Messages that aid in debugging
These are the main features that we will enable in the next quarter
- Correlation - AutoRCA
- KPI Import
- Metabase (partially done)
- Superset
- Looker
- Dbt
- API support for KPI management & analysis
- Anomaly detection
- Investigate Anomaly List
- More granular runtime frequency - currently 1 day
- Model/Params change
- Data quality
- Deeper DQ metrics which account for data distribution
- Ability to define arbitrary DQ metrics with Great Expectations
- KPI definition catalog
- Single data source
- Multi data source
- Forecast as input for RCA
- K8 configuration for horizontal scaling
We will enabling the following new connectors:
- S3
- Data lake support including Delta Lake
- Google Playstore
- RPA based connectors platform (longer term)
We are working on the following to improve & support data at larger scale:
- Data scalability
- Distributed Pandas support - Koalas, Dask
- Compressed Data/ Metrics Store (longer term)
- Interactive analysis at scale
- Implementation of Pinot & Druid based data store to enable interactive analysis
- ML scalability
- Model warm start where possible for heavier models
- Feature & model store