Photo: Alex SG – shutterstock.com
IBM has acquired Israeli startup Databand.ai. The software house, founded in 2018, has developed a so-called observability suite. The software for data monitoring is intended to help companies to identify and correct problems with their data as quickly and automatically as possible. For example, this should prevent erroneous data from corrupting machine learning and AI workloads and leading management to make bad decisions with misleading results.
With the acquisition, IBM complements its software portfolio in the areas of data, AI and automation. Databand.ai is IBM’s fifth acquisition this year. Since Arvind Krishna took over the helm of the IT pioneer as CEO in April 2020, more than two dozen IT companies have been bought.
The employees of Databand.ai, headquartered in Tel Aviv, are to be merged into IBM’s Data and AI division and will continue to expand the portfolio there, including Watson functions and IBM Cloud Pak for Data. The companies did not disclose financial details of the transaction. The acquisition was already completed on June 27, 2022.
In view of the rapidly growing amounts of data, companies have difficulties in correctly assessing the condition and quality of their data, IBM said. Therefore, software for data observation is increasingly important for data teams and data scientists in companies. According to IBM, the aim is to better understand the status of the data in the IT systems and to identify and correct problems such as anomalies, faulty data changes or incorrectly configured data pipelines automatically and, if possible, in real time. According to Gartner, poor data quality costs each company an average of $12.9 million per year.
Photo: drserg – shutterstock.com
The Databand.ai solution uses historical trends to calculate statistics about data workloads and data pipelines directly at the source. This should make it possible to determine whether the data quality is sufficient for ML and AI requirements or where problems could arise. The platform can be operated as a SaaS solution and, in the enterprise version, can also be installed in your own data center. Various sources can be connected to the data radar, from data lakes to streaming platforms to classic database and data warehouse solutions.
What a data scientist must be able to do
IBM apparently plans to link Databand.ai to its IBM Observability by Instana APM solution. For example, Databand.ai capabilities could alert data analysts when data they are using for an analytics system is incomplete or even missing. Since this data usually comes from business applications, Instana can then help users in the next step to clarify where the faulty data comes from or why an application service is delivering bad data or no data at all. Together, Databand.ai and IBM Instana would provide a more complete and better explainable view of the entire application infrastructure and data platform system, IBM promises.
“Data-driven companies depend on high-quality, trusted data to drive their mission-critical processes,” said Daniel Hernandez, general manager of data and AI at IBM. With Databand.ai, IBM offers the necessary observability functions to provide trustworthy data for AI at scale.
According to Josh Benamram, co-founder and CEO of Databand.ai, not being able to see that their data platform is performing ineffectively or even malfunctioning is a big problem for companies. It’s about avoiding nasty data surprises by identifying and fixing problems before they have a costly impact on the business.
#IBM #buys #Databandai #data #quality #machine #learning
