Highlights:

  • The company mentioned that the AI model adapts to usage patterns to tailor alerts and reports.
  • The assistant can also grasp patterns of cost consumption, assisting companies in improving control over their cloud expenses and predicting consumption based on learned behavior.

Acceldata Inc., a data observability startup, has introduced artificial intelligence features to its platform, including a “copilot.” This feature empowers DataOps teams to customize AI assistants according to their technology and business landscape.

This encompasses the capability to establish safeguards, ensuring that business context, regulatory requirements, and human oversight are integral aspects of the equation.

Utilizing technology from its acquisition last fall of Bewgle Inc., a developer of unstructured data analysis software, Acceldata’s copilot examines and produces alerts regarding deviations in data freshness, data profiling, and data quality, thereby enhancing the reliability of the data. With large language model technology, it improves communication between technical and business stakeholders by automatically producing human-readable descriptions of data assets, policies, and rules.

Configured with retrieval-augmented generation, the assistant can ascertain “what data sets are being used, what is being included in reports and what is not included,” said Co-founder and Chief Executive of Acceldata, Rohit Choudhary. “People often don’t know where their data is being used. This tells you about the data that’s most important to you.”

Learned Behavior

The company stated that the AI model comprehends patterns in data usage to tailor alerts and reports. For instance, the deviation from the usual timing of a report’s arrival or alterations in the demographic profile of a customer cohort can signal data drift. Choudhary said, “All of these are now being done automatically, not figuring out after outcomes have been seen.”

The assistant can also grasp cost consumption patterns, aiding companies in enhancing control over their cloud expenses and predicting consumption through learned behavior. In the context of cloud data warehousing, “the user community never had cost considerations before, but those costs add up because Snowflake is consumption-based. We can find out which SQL queries are poorly formed with machine analysis instead of scrounging through lines of SQL.” mentioned Choudhary, citing a well-known cloud data warehousing vendor. Acceldata also assesses intermediate data assets, notebooks, and programmatic access.

The features of LLM simplify the policy creation process and minimize the risk of errors. Choudhary said, “We’re looking at the pattern of data usage to learn how different data sets are being put together. An employment contract has a set of fields and documents that you should look at. Any data set that represents that data can automatically be verified. If somebody has made a wrong entry into an upstream employment database, we can automatically discover that it is not following the pattern without creating a manual rule. We then have the ability to create a policy.”

Acceldata secured over USD 105 million in funding from various notable investors. In late 2022, the company open-sourced its platform and generated revenue by selling exclusive extensions tailored for enterprise customers.