Starburst has unveiled its latest feature, the AI Data Assistant (AIDA), designed to transform static reporting into real-time, context-rich decision-making.
This new capability allows users to interactively query and analyse enterprise data using natural language, thus streamlining the process of converting queries into tangible insights. By addressing common challenges such as the long wait times for dashboard creation and the reliability of data, AIDA aims to improve the speed and confidence in decision-making processes.
Enabling governed data access
Enterprises have struggled with accessing dispersed data effectively, often mistaking centralised systems for true transformation. AIDA offers a solution by providing controlled, immediate access to verified enterprise data, facilitating faster and more informed decisions. This evolution from static data handling to dynamic data interaction represents a significant shift for businesses striving to leverage their data assets effectively.
AIDA offers a solution by providing controlled, immediate access to verified enterprise data
AIDA introduces advanced reasoning through the ReAct framework, which leverages live data sampling and metadata analysis to produce insightful responses akin to those of a data analyst. Additionally, AIDA customises outputs based on user roles, catering to different needs with comprehensive technical details or succinct summaries. Organisations can also seamlessly integrate their branding with AIDA, utilising it without further development, currently available within the Starburst Enterprise Platform (SEP).
Diverse support and integrations
SEP supports a range of large language models from Anthropic, OpenAI, and AWS Bedrock, allowing companies to select technologies based on their unique security and cost requirements without being tied to a single vendor. Upcoming enhancements, such as AIDA Studio and the AIDA MCP Client, are expected to extend integration capabilities with external systems and promote a broader use of AIDA for automating essential enterprise tasks.
A new governance layer adds configurable control over AI interactions, ensuring compliance with organisational policies and data protection regulations. This layer prevents sensitive data exposure and restricts discussions on delicate topics. Starburst’s approach contrasts with traditional methods that centralise data; instead, it enables AI to function directly on decentralised data systems, avoiding the need for data relocation.
AI-Driven Business Intelligence
Starburst supports a federated context layer that aligns business policies and definitions
“Most companies are still approaching AI the wrong way, focusing on models instead of the data those models depend on,” remarked Justin Borgman, Starburst’s Co-founder and CEO.
He highlighted that AIDA addresses the challenge of integrating AI into business decisions without relocating data or compromising its governance. By enabling AI to connect with distributed data sources directly, Starburst supports a federated context layer that aligns business policies and definitions.
Supporting this perspective, Kevin Petrie from BARC US noted that Starburst facilitates analytics democratisation by providing governed data access. This empowers diverse stakeholders and enhances their decision-making by enabling intent- and persona-specific reasoning. Whether it is recovering lost revenue by pinpointing billing discrepancies or reducing manual investigation time through automated fraud and compliance checks, AIDA positions enterprises to utilise their data strategically and efficiently.
Starburst, a pioneer enterprise intelligence platform, announces its AI Data Assistant (AIDA), a new capability that helps organisations move from static reporting to faster, more context-aware decision-making. With AIDA, users can explore and analyse trusted enterprise data using natural language, making it easier to turn questions into actionable insight.
Teams wait months for the creation of dashboards, export the results into spreadsheets for further analysis, and still question whether the numbers can be trusted. That gap makes it difficult to act on data when it matters most.
Trusted enterprise data
Users, applications, and AI systems need governed access to data across the business to act with speed and context. Yet for years, centralisation has been treated as transformation, even in enterprises where data is spread across clouds, platforms, and operational systems.
With AIDA, organisations can move beyond static reporting and give users governed, on-demand access to trusted enterprise data, enabling faster, more context-aware decisions.
What’s new in AIDA
- Advanced Reasoning Capabilities: AIDA leverages a ReAct (reason–act–observe) framework to move beyond query generation into true analytical reasoning, combining live data sampling and metadata analysis to reach a well-grounded answer. The result is an assistant that reasons through problems like an analyst, not just a text-to-query translator.
- Persona-Based Outputs: AIDA tailors responses based on user role, delivering detailed technical explanations for data practitioners and concise, decision-ready summaries for business leaders.
- White Labeling: Organisations can apply their own branding to AIDA to create a seamless internal analytics experience without additional development. Available today in Starburst Enterprise Platform (SEP).
- Flexible LLM Support: Within SEP, AIDA supports multiple LLMs, including models from Anthropic, OpenAI, and AWS Bedrock, enabling enterprises to choose the model that best fits their technical, security, and cost requirements without vendor lock-in.
Important enterprise tasks
Coming in Q2, Starburst plans to release the following:
- AIDA Studio: An extensibility layer that enables integration with external systems, incorporation of unstructured business context, and creation of custom skills to orchestrate workflows across tools like Slack, Jira, and Google Workspace.
- AIDA MCP Client: The AIDA MCP Client Layer gives AIDA the ability to interact with and pull context from enterprise applications such as Slack, Jira, GitHub, using the open Model Context Protocol (MCP). Users can add context to inform AIDA’s outputs, and even use AIDA more broadly as an automation hub for important enterprise tasks.
- Guardrails: A configurable governance layer that controls AI interactions and outputs, enforcing policies beyond underlying data access. Organisations can restrict sensitive topics and prevent exposure of personal data, ensuring safe and compliant AI usage.
Centralised data architectures
"Most companies are still approaching AI the wrong way, focusing on models instead of the data those models depend on," said Justin Borgman, Co-founder and CEO of Starburst. "The real challenge is applying AI to business decisions without moving data or compromising governance. Starburst’s AI Data Assistant is built to solve that by providing access to trusted, distributed data from across the enterprise."
Unlike traditional approaches that depend on centralised data architectures built for business intelligence, Starburst enables AI to operate directly on distributed data across lakes, warehouses, cloud object storage, and operational systems. While competing approaches often require data to be moved into a vendor-controlled environment before AI can act on it, Starburst brings AI to the data wherever it resides without lock-in.
AI connecting distributed data
By applying governance, definitions, and access controls consistently across data sources, Starburst provides the federated context layer required for enterprise AI connecting distributed data, business meaning, and policy into a single AI-ready foundation. AIDA is the interface. The Starburst platform is what enables it to operate across the enterprise.
“As enterprises seek to democratise analytics with agentic AI, they need governed access to distributed datasets,” said Kevin Petrie, Vice President of Research at BARC US. "Starburst meets this requirement and goes further to enable intent- and persona-specific reasoning on federated inputs. This helps diverse stakeholders make smarter decisions in the context of the business.”
Identifying billing discrepancies
Recover lost revenue by identifying billing discrepancies across contracts, usage, and invoicing data, quantifying the impact, and triggering corrective actions, helping reduce the 1–3% of revenue often lost to leakage.
Prevent customer churn by detecting early warning signs from usage, support, and sentiment data, generating customer health insights, and prompting timely interventions, turning at-risk renewals into retention opportunities.
Accelerate fraud and compliance investigations by surfacing suspicious activity across transactions and customer data, enriching it with full context, and automating case creation — reducing manual investigation time while improving accuracy.
