Big Data Security Analytics , Next-Generation Technologies & Secure Development
IBM Buys Startup Databand.ai to Address Data Quality Issues
Databand.ai Ensures Data Is Being Accessed by the Right Users at the Right TimeIBM has purchased a data observability startup to help organizations address data errors, pipeline failures and poor quality before it affects their bottom line.
See Also: The Role of the Data Storage Platform in Providing Data Resiliency
The Armonk, New York-based technology giant says its acquisition of Tel Aviv, Israel-based Databand.ai will help businesses ensure that trustworthy data is being put into the hands of the right users at the right time. Databand.ai can alert data teams and engineers when the data they are using to fuel an analytics system is incomplete or missing, according to IBM.
"Our clients are data-driven enterprises who rely on high-quality, trustworthy data to power their mission-critical processes," IBM Data and AI General Manager Daniel Hernandez says in a statement. "When they don't have access to the data they need in any given moment, their business can grind to a halt."
Terms of the acquisition - which closed June 27 and was announced Wednesday - weren't disclosed, though Israeli publication Globes reported that IBM paid $150 million for Databand.ai. An IBM spokesperson declined to comment on the purchase price reported by Globes. IBM's stock was up $0.46 - 0.33% - Wednesday to $138.08 per share.
Former Sisense Product Manager Josh Benamram founded Databand.ai in 2018 and has led the company since its inception. Databand.ai currently employs 51 people and in December 2020 closed a $14.5 million Series A funding round led by venture capital firm Accel, according to LinkedIn and Crunchbase (see: IBM to Buy Red Hat for $34 Billion).
"You can't protect what you can't see, and when the data platform is ineffective, everyone is impacted - including customers," Benamram says in a statement. "Joining IBM will help us scale our software and significantly accelerate our ability to meet the evolving needs of enterprise clients."
What Does Databand.ai Do for IBM?
Becoming part of IBM will allow Databand.ai to fortify its observability muscle for broader integrations across more of the open-source and commercial offerings powering the modern data stack. Businesses will have the flexibility to run Databand.ai either as-a-service or as a self-hosted software subscription, according to IBM.
Databand.ai builds on IBM's existing Instana platform to provide a more complete and explainable view of the entire application infrastructure and data platform system. This can help organizations prevent lost revenue and reputation, according to IBM.
"Data observability takes traditional data operations to the next level by using historical trends to compute statistics about data workloads and data pipelines directly at the source, determining if they are working, and pinpointing where any problems may exist," Mike Gilfix, IBM's vice president of data and AI product management, writes in a blog post.
Gilfix says Databand.ai collects data pipeline metadata across key technology in the modern data stack and uses that to build historical baselines for data pipeline behavior. This allows Databand.ai to detect and generate alerts on anomalies while the data pipelines run as well as resolve anomalies in an automated fashion without affecting the delivery, according to Gilfix.
Databand.ai's technology will improve mean time to discovery by detecting and executing on data pipeline issues in real time instead of reacting afterward, Gilfix says. Mean time to repair will also improve since the contextual metadata provided by Databand.ai helps data engineers focus on the source of the problem rather than debugging where the issue stems from, according to Gilfix.
IBM will make Databand.ai's data observability capability available on a stand-alone basis, but Gilfix recommends using it alongside the company's multi-cloud data integration, data governance and privacy, and trustworthy AI capabilities to more effectively automate the data life cycle. Databand.ai will integrate with these other use cases for improved results where both are applied, according to Gilfix.
"Catching data quality problems at the source helps enable the delivery of more reliable data," Gilfix writes in the blog post. "Monitoring both static and in motion pipelines while delivering high quality metadata enables a faster time to value than would otherwise be possible."
How Can Data Observability Benefit Security?
Data observability platforms historically have been used to ensure an organization's data is clean and useful for downstream teams by probing for issues such as data lineage and algorithmic drift, Forrester Vice President and Principal Analyst Jeff Pollard tells ISMG. But security leaders are increasingly looking to get visibility into their organization's data to ensure that it isn't being tampered with by adversaries, he says.
Security and data leaders were each erroneously assuming that the other team was responsible for ensuring that data hadn't been tampered with, but Pollard says it's increasingly clear that CISOs have a role to play when it comes to data integrity. Pollard expects this to become a greater area of focus as businesses work to ensure adversaries aren't tampering with or poisoning their data.
"The more you automate things, the more you're relying on sensor data and input and things like that," Pollard says. "They've got motivation to have you make the wrong choice."
Investment activity around data observability has picked up over the past two to four years to ensure the integrity of the data that's increasingly being used in artificial intelligence and machine-learning algorithms, Pollard says. Data scientists typically understand what's being input into the algorithms, but Pollard says visibility and telemetry for those who aren't practitioners is limited due to closed systems.
Security teams were traditionally focused on understanding and correlating log data, but observability platforms tend to be much closer to the applications and data actually in question, according to Pollard. He therefore encourages security leaders to build programs, processes and practices around app and data observability pipelines in better alignment with practices across the organization.
"Security leaders need to keep their eyes focused on concepts around observability because this will be an area where they are starting to make sizable investments in the future," Pollard says. "And it's very much a change from what they've done in the past."