Release Overview: March 2021¶
Aunalytics is excited to announce the March 2021 Daybreak release to our clients. This release will provide clients with model and site enhancement information along with any fixes to existing functionality we have included.
Natural Language Answers¶
Daybreak exists to provide answers to the questions users of all backgrounds have about their organizations' data. Now, thanks to months of research and development, Aunalytics is excited to debut a new tool that puts data within the reach of everyone: natural language answers. Natural language answers is a machine learning service that enable users to interact with their industry intelligent datamarts by simply asking questions.
Natural language answers integrates into the Daybreak web app to provide users a way to get answers from their datamarts as easily as typing a question in the Google search bar. Behind the scenes, Aunalytics' proprietary NL2SQL technology translates these questions into structured queries to deliver datasets that respond to questions a user might ask.
Aunalytics' Innovation Lab data scientists have created a machine learning algorithm to understand the questions people ask about their data in a particular focus industry so that it can translate those requests into structured queries suitable for running against a SQL-compliant datamart.
Thanks to these efforts, our clients using Daybreak for the financial services industry can now begin the day with answers by asking Daybreak questions like "give me a list of mortgages opened last quarter" or "Show me CD accounts with a maturity date in the next month."
Learn more about this exciting new feature and how to use it here.
Daybreak Insights provide a new way to extract value from data by allowing the creation of dashboards of visualizations from the results of their datamart queries. With this release, users will be able to access the "Insights" tab in the Data Builder's results area (at the bottom of the Data Builder tool). The Insight's tab provides a dashboard where users can create, save, and share charts of their query's results.
Currently, four types of visualizations are supported: numerical summaries, column charts for data points grouped by category, line charts for data points grouped by time interval, and donut charts to show proportional distribution by category.
Learn more about Insights and start creating them today here.
Datamarts are new analytics cloud management components for the Aunsight platform designed to support the needs of our Daybreak industry intelligent datamarts and web app. Datamart objects provide a single point of configuration for defining the structure and schema for the data that will be exposed in a Daybreak datamart. This new component allows solutions engineers to define the way data will be presented in a datamart and enable other components such as Workflows and Connectors to copy and transform data into it according to the rules defined for it.
Solutions engineers can now begin to implement datamarts to facilitate the migration of datasets into the Aunsight high performance, SQL-compliant storage infrastructure. Datamart configuration settings allow engineers to specify transformations to apply to the source datasets during migrating into the storage engine, and how that data should be presented to the user in the Daybreak web app. As a result, Datamarts provide a management surface to streamline the creation and deployment of Daybreak datamarts every time a datamart is refreshed.
Users whose Daybreak implementations make use of Datamarts may immediately notice significant improvements in query performance as their data now resides within our high-performance SQL-compliant storage engine. Less immediately noticeable but equally important, enhanced capabilities with data migration will occasion fewer interruptions in the nightly pipeline jobs that refresh Daybreak datamarts. It should be noted that existing clients will not be able to take advantage of these features until Datamarts are implemented as a part of their solution pipeline.
Pipelines are new Aunsight automation components for managing the delivery of client solutions. Similar to workflows, pipelines define a sequence of workloads that should be run in order to process data for a client solution, but unlike workflows which are often created to orchestrate small processes that form part of the overall production solution, a pipeline defines the complete, end-to-end process and provides mechanisms to manage failures in the stages used to deliver finished datasets to clients via a solution like Daybreak.
Pipeline objects allow solution engineers to define the stages of a production solution and manage related components that comprise those stages, such as Workflows and AuQL scripts. Pipelines also streamline the parameterization of component stages by passing certain metadata such as storage and compute resources to use onto the dependent stages of a pipeline. Finally, Pipelines enable more resilient management of production workloads through retry mechanisms that can ensure a client solution can still be delivered on schedule even if one component experiences a job failure.
|AUN-14610||Datamarts: Add permissions to policies and standard roles|
|AUN-14580||Add new "log" category of operations to lib-pig|
|AUN-14573||add kill button to datamarts job page|
|AUN-14556||Daybreak: Add row count to data results|
|AUN-14541||Remove references to 'rowDelimiter' in atlas records|
|AUN-14540||Add metadata field in the UI to show started_at field for datalab|
|AUN-14539||Model service: Disable deletion for published model versions|
|AUN-14526||Daybreak: Rename headline and button label on NLP landing page|
|AUN-14485||Edit atlas record directly from UI|
|AUN-14400||Daybreak: Add tooltips table fields in SQL editor|
|AUN-14379||Daybreak: Add one-to-many to query wizard|
|AUN-14213||DSV output format|
|AUN-14028||Rename cancel to close in expression builder|
|AUN-14027||Improve validation in expression builder|
|AUN-14639||Add index(s) to prevent error when paging through long Dataflow lists|
|AUN-13908||memento series dependency timestamp mode|
|AUN-14367||Add AuQL job kill method and command to lib-au-js and toolbelt|
|DATAINT-480||AU TXWF Date format standardization (8601)|
|DATAINT-38||Aon Mongo Accessibility|
|PZ-387||Improve Warehouse Workflow to better handle failure|
|AUN-14708||Bug in relation column showing in wizard|
|AUN-14699||Fixing the dataflow builder bug for not adding extra input fields for certain operations|
|AUN-14681||Datamart relax table compatibility checking|
|AUN-14659||Daybreak - Insights -numeric aggregators are listed on non-numeric fields|
|AUN-14642||The ingest tab of Datasets is blank in the mar21 release candidate|
|AUN-14641||Daybreak-Insights: Remove Insights from all SQL views|
|AUN-14578||Custom pipeline job names get overwritten with pipeline name during pipeline execution|
|AUN-14548||Workflow stats not loading|
|AUN-14564||Daybreak: Fix issue with query list open/close icons|
|AUN-14497||Moment dependency is taking up significant space in lib ui|
|AUN-14432||The SQL Read Connector allows spaces with field names, Aunsight doesn't, need schema creator to address that.|
|AUN-14386||Check-context-permissions throws error when checking only global perms|
|AUN-14337||Daybreak-Page doesn't scroll when page has a lot of content|
|AUN-14329||Deleting a memento series with the same key twice throws an error|
|AUN-14328||Update Storybook version and fix modal bug|
|AUN-13962||Query Workflow component has small bug in the failure return.|
|AUN-4116||Injecting pig code is possible via dataflow builder|
|AUN-14637||Fix sonar issues|
|DATAINT-456||Naveego Agent leaving rogue plugin processes running|