The purpose of Keboola Connection is to provide an agile and affordable data integration framework.
Keboola Connection is a hybrid cloud-based set of integrated tools which supports the whole data workflow process, from the point of data extraction, preparation, cleaning, warehousing and all the way to its integration, enrichment and loading.
The basic framework is provided by Keboola and can be extended with third party components at all stages of the process – the best example being Machine Learning or Data augmentation component (Weather, Geo, Basket Analyses, etc.) applications – and seamlessly integrated within one environment.
With Keboola you can instantly extract data from more than 170,000 different sources with no programming necessary.
• More than 70 pre-built custom data connectors, ranging from marketing to finance applications, with UI. Clients with development resources can also build their own.
• REST API data connector, configurable for thousands more data sources in a matter of mere hours!
• Ability to flexibly pull all data from required sources and select metrics; no one-size-fits-all solutions.
Storage API receives data from data connectors and stores it securely for further operations.
• The data can then be accessible for transformation, such as data blending, mashing up and augmentation, via direct access into Storage API with a valid token.
• Internal tasks are conducted in isolated environment for security reasons and all of their actions are being logged for audit purposes.
This log is visible and searchable to all users of the project and offers full rollback function.
• KBC uses engines best suited for the nature of each task, such as MySQL, Redshift or Snowflake.
• Sandbox is instantly available to you with a copy of data, so you can experiment anytime you want without worrying about losing the core data.
• Transformations can be re-used and shared among projects.
• We hate limitations and believe in customer’s freedom to choose the right consumption point for each project and use case. Keboola is agnostic on the input as well as on the output. Some of the most popular data loads currently supported are GoodData, Tableau, YellowFin, Redshift, Oracle database, Dropbox, Amazon S3, MySQL, ElasticSearch and Google Drive.
• “Full circle” concept. We believe that a new value is not always created by loading data to dashboard. In some use case, for example advanced segmentation of customer base, it is useful for creating a new mailing list or custom audience in marketing automation tools, so we also let you send your enriched data there. Like with Data Connectors, there is no limitation to where you can send your data.
• Once data flow is set up, Orchestrations take care of running the whole sequence – automatic data updating, transforming and loading.
• Occasional hiccups like credential expirations are easy to spot.
• Jobs keep track of all tasks and help oversee performance of all components. It means that transparent security mechanisms are in place throughout the whole process.
• Use pre-built apps to help perform various types of analysis, from data type assistant finder and anomaly detection to basket analysis or churn prediction, free of charge. They can be released with only a couple of clicks.
• We can prepare custom algorithms for you. You can even use your own solutions, thanks to the docker image’s help.
• Privately deploy your own scripts.
• You can also use third party services, such as Geneea available in Keboola, to leverage Natural Language Processing and machine learning algorithms on your unstructured data.