Elixir Ambience

Getting Started

Ambience is a suite of tools providing services primarily for business intelligence (BI) purposes. Data visualization, reports and ETL are the most common usage of Ambience.

Administration

To start using Ambience, ensure that you have access (authentication) to the system. Depending on the set up, this can be done through either:

the Identities module if authentication is handled internally within Ambience, which is the simplest set up, or
externally (e.g. external LDAP server)

Access to each Ambience module interface and some module features are controlled by privileges and roles assigned to the user. By default, there is a system administrator account available upon set up. This account is already authenticated and is authorized to access all administration modules:

To add new users to Ambience, ensure that the new users is added through the Identities module first then through the Users module, using exactly the same name. Refer here for a more comprehensive guide on the Users module. Assigning roles can be done through the Users module and through the Roles module as well.

Datasets

Data is the most important component in any BI solution. To start, add a new dataset through the Datasets module.

There are two main types of datasets:

MongoDB datasets, which are directly read from MongoDB (normally used as staging or storage after a performing ETL on the actual data source)
ETL datasets, which are sourced from the output from a specific ETL chain

Most views on the Dashboards module (data visualization) require using datasets to render. Ideally, ETL should be done on the data source and stored into MongoDB so MongoDB datasets can be used.

Datasets, ETL and Dashboards modules work hand-in-hand to produce BI solutions.

Refer here for a more comprehensive guide about the Datasets module.

ETL

The ETL module provides many steps and services with specific purposes. There are steps that trigger SMTP services, read XLSX files, write files, load files, manipulate loaded data and many more.

Though the ETL module is a complex and powerful module that caters to more advanced technical users, the average user would still find the need for it from time to time, for instance, to make simple transformations on data loaded in a MongoDB instance used by Ambience. Other commonly used steps are:

joins (i.e. equivalent of table joins in relational databases),
report generation (i.e. running the report engine),
field manipulation (e.g. date format manipulation, JSON string manipulation)
chain calls (i.e. running another chain’s, using another chain’s results)
MongoDB read and write steps
exposing the results of the chain as a dataset that can be used in the Datasets module

Refer here for a more comprehensive guide about the ETL module.

Imports

The Imports module provides the most straightforward way to load data into any MongoDB instance that Ambience has write access to. Simply prepare a file (e.g. XLSX, CSV) and import it into a MongoDB collection. The collection (in the default and recommended config) is easily accessible through the Ambience modules, such as ETL, Datasets and Dashboards modules.

This module makes it easy to quickly load and use data in the said modules. Refer here for a more comprehensive guide about the Imports module.

Uploads

Uploading files can be done through several ways. If the purpose is to upload a file into Ambience (e.g. HTML file, PDF file) and make it accessible for purposes such as adding a file download link on a dashboard, displaying an HTML file on an iFrame on a dashboard, making it readable from the ETL module, using the Uploads module would be the simplest way to do this.

Another option is to manually copy the files into a directory (e.g. shared directory, ) accessible to Ambience.

Refer here for a more comprehensive guide about the Uploads module.

Dashboards

The Dashboards module provides many ways to visualize data. To start, simply create a dashboard. This provides a thorough guide on how to add a new dashboard.

Upon creating a new dashboard, it is immediately loaded. With the appropriate privileges (access rights) and ownership of the dashboard, the owner can configure the dashboard template using the Designer. Guide on using the Dashboard Designer can be found here.

Each dashboard can have several pages (or tabs) and each page can contain several views. Most view types require datasets to be identified to be able to present data. These view types include charts such as the pie chart, bar chart, line charts and many more.

The pie chart is one of the most straight forward views. Simply set the “Dataset” to an existing dataset to identify the data to be displayed, set the “Key” property to a field on the dataset that would be used to group the data then set the “Value” property to a field on the same dataset that would be used to allocate the portion sizes. Aggregation is set to “count” as default so only the three properties mentioned are really required to render a pie chart.

There are over 250 different view properties, some shared across a few view types, some are unique to a certain view type.

Clicking on the “Save” button on the Designer is required to ensure changes in preview are put into effect (i.e. when another user views the dashboard).

There are also some views that provide high levels of customization, such as the HTML view. These are suitable for more technical users though as they do require some scripting.