ATACMA launched new AI-powered features on Thursday as part of its latest platform update that provides natural language explanations for data lineage characteristics.
Data systems are a way to ensure data quality. By tracking and documenting data, data can be tracked and documented through systems that are converted and prepared from sources for analysis, stored and replicated to be accessible to analytics and AI tools, and moved between systems, allowing organizations to see if the data is reliable and high quality.
Ataccama first added AI-powered data lineage to its February platform update, providing an automated way for users to track data flow and transformation.
However, since SQL or Python coding skills are required to read and interpret the output generated by AI, customers needed technical experts and reasons why the issue was flagged to help business users understand the systematic metrics of their data. This has slowed reviews and sometimes prevented business users from acting on the data in a timely manner.
With the launch of Ataccama One version 16.2, vendors will add AI-generated natural language descriptions to allow technical users to understand the quality of data they are using and make decisions without waiting for the help of trained professionals.
This is an important addition for Ataccama users. Our research shows that data quality is a top use case for agent data management, and Lineage helps to ensure data quality.
Kevin PetryAnalyst, Barc Us
According to BARC US analyst Kevin Petrie, it's a valuable addition to Ataccama's platform, given the new features make data lineage more transparent.
“This is an important addition for Ataccama users,” he said. “Data quality is a top use case for agent data management, and research shows that Lineage helps ensure the quality of your data.”
Based in Toronto, Ataccama is a data management vendor focused on data quality. In addition to data lineages, Ataccama offers master data management, data governance and data observation capabilities.
New Features
Since the launch of CHATGPT in November 2022, companies have been increasingly investing in AI projects, representing a significant improvement in Generated AI (Genai) technology, making data quality more important.
While always essential for the success of analysis and AI projects, there was more human involvement in decision making prior to the spread of Genai and the recent advent of fully autonomous agent AI applications. Human involvement prevents decisions based on bad data. However, AI tools can be entrusted with acting independently, reducing and sometimes eliminating failsafes for human surveillance.
As a result, the data that notifies Genai and agent applications must be of higher quality than ever before.
Ataccama's new AI-powered data lineage features provide a natural language explanation of how data was transformed throughout the lifecycle. As part of this, we will discuss filters, joins, and calculations so that business users can understand the logic behind the description.
According to Matt Aslett, an analyst at ISG Software Research, the latest Ataccama updates are designed to make data easier to understand and trust users, adding valuable features.
“Ataccama has greatly enhanced the systematic capabilities of data, allowing users to better understand the properties of data assets and how they use them across their organization,” he said. “The new capabilities will allow business users and data readers to understand the system of data assets without the need for deep technical expertise.”
Meanwhile, as companies are investing more in AI development, transparent representations of data lines support AI and machine learning governance by showing how data lines' inputs (the original intake points of data) relate to users, pipelines, and model outputs.
“This will make the AI architecture more transparent and the models are easier to explain to key stakeholders, such as customers and auditors,” he said.
According to Jessica Smith, Vice President of Data Quality at Atacama, the feedback provided some of the driving forces for adding natural language descriptions for data lineage calculations, requiring that customers have the same data lineage information as technical experts.
“Customers told us that these insights are needed to ensure business users have the same access — clear, fast, without technical help,” she said. “It's about removing friction for non-technical teams who need to understand data quickly and confidently.”
Beyond AI-powered natural language explanations for data system characteristics, Ataccama's platform update includes:
A systematic diagram that provides a high-level view of data flows with details that users can drill on demand.
A secure lineage that allows for metadata extraction from on-premises and restricted environments without forcing users to move sensitive information to the cloud.
Connecting to Google BigQuery and Microsoft Azure Synapse allows data profiling and data quality workloads to run without uploading to Atacma. This saves organizations on data exit costs.
According to Aslett, perhaps the most important systemic capacity is the safe system.
“The ability to extract metadata from sensitive data is an important feature that allows users to better understand the data environment while still maintaining data sovereignty regulations and policy compliance,” he said.
Additionally, Aslet emphasized pushing workloads down into the data warehouse.
Looking ahead
According to Smith, vendors will continue to focus on using AI to promote data quality and create trust in data with the latest Ataccama updates that are now available in general.
In particular, Ataccama aims to add a standard that includes new agent AI features (a protocol for how agents interact with other tools and systems) that include input from the Model Context Protocol server.
Additionally, Ataccama is considering adding new capabilities to address unstructured data and reference data management, and expanding the data observability capabilities, Smith continued.
“We are exploring new features… to provide our customers with deeper insights and control over their data environment,” she said.
Meanwhile, Petrie proposed that Ataccama partner with more AI and machine learning providers to allow ATACCAMA functionality to integrate with AI tools to improve the performance of AI/ML models and applications.
“[Ataccama’s] “The rich lineage metadata helps AI/ML platform vendors manage data along with AI/ML models and agent applications, contributing to an end-to-end view of input, logic and output,” he said.
Eric Avidon is a senior news writer at Informa TechTarget and a journalist with over 25 years of experience. He covers analytics and data management.
