Data Therapist Hour vol.2. Yuliia Tkachova, CEO and co-founder @ Masthead, hosting it with Max Ostapenko, a Data Product and Analyst Team Lead @ idealo. Get answers What is Data SLA? What are the components of Service Level Agreements? Data SLA must-haves and its mistakes.
The video starts with an introduction to data SLAs. A data SLA is a public promise to deliver a specific level of data quality. There are two main reasons why a company might need a data SLA. First, if a company’s product or service relies on data, then the company’s service level agreement (SLA) essentially is the same as the data SLA. Second, even if a company’s product or service does not rely on data, a data SLA can be useful if the company’s teams rely on data to make decisions.
According to the video, there are three main components of a data SLA: service level indicators (SLIs), service level objectives (SLOs), and service level agreements (SLAs). SLIs are specific metrics that measure how well a data service is performing. For example, an SLI could be the percentage of time that a data pipeline is operational. SLOs set targets for SLIs. For example, an SLO could be that a data pipeline must be operational 99% of the time. Finally, SLAs are agreements between data providers and data consumers that outline the SLIs and SLOs that will be used to measure the performance of a data service.
The video also discusses some of the common mistakes that companies make when implementing data SLAs. One common mistake is to set SLOs that are too ambitious. Another common mistake is not to communicate SLAs effectively to data consumers.
The video concludes with a discussion of some of the tools that can be used to help companies implement data SLAs. One tool that is mentioned in the video is Masthead Data, a data observability platform that can help companies monitor the health of their data pipelines.
