For decades, the answer to every meaningful business question has followed the same structure. Vital operational inquiries were transformed into queries, results were visualized on charts, dashboards were created from these charts, and insights were derived.
This approach was always, in some sense, problematic.
It left key stakeholders with a backward-looking view of the organization rather than a clear map for the future. The entire methodology was static and slow-moving, and while the approach provided insights, they were inherently limited. Most of all, they did not provide a dynamic, interactive, conversational approach centered on human-level questions. In contrast, dashboards remain only part of the question-to-data-to-decision workflow. There is an element of interpretation, recommendation, and action that dashboards have never covered.
It was the best anyone could do, but that is not the same as the best possible solution.
Starburst AIDA now Generally Available (GA)
AI has changed this approach entirely, and Starburst’s AIDA is at the forefront of that change. Today, as we announce the general availability (GA) of AIDA, we are breaking that cycle by introducing a solution that fundamentally changes how enterprises interact with data. AIDA is far more than an AI agent. It is built to address the problems with past dashboards, replacing backward-looking information with forward-looking insights. It replaces the static, rigid dashboards with a dynamic, conversational interaction model and underscores it all with the trust that only Starburst can provide through its unique access to the context layer and semantic layer needed to make enterprise AI decisions accurate and governed. It is designed for a new era, one beyond BI and beyond dashboards.
For anyone not familiar with Starburst’s AIDA, check out this video for more information on the core problem it is designed to solve.
Overcoming traditional limitations of Business Intelligence (BI)
It’s worth revisiting the core problem AIDA is designed to solve. And for that, you need to look at the failings of the static dashboards it is designed to replace.
Static vs dynamic, rigid vs human
Traditional dashboards are fundamentally designed to surface a number. They aggregate, filter, and render data to deliver a historical view of what has already happened. However, this output is only a fragment of the complete business workflow, not the entire workflow. Once a dashboard is delivered, the most critical work still falls entirely on human shoulders.
The center of this problem is a human one. BI dashboards, by their nature, are not centered around humans in the right way. In BI workloads, each question needs to be thought of in advance. Follow-up questions become difficult because the timeframe involved in receiving answers is not suited to fast, iterative repetition.
Instead of adapting to how people actually think and work, traditional tools force users to adapt to the rigidity of the database. A person still has to understand what the numbers mean in context, propose a path forward, build alignment with stakeholders, and execute the final decision. This manual friction is precisely where organizational velocity vanishes into weeks of traditional data work. It is a foundational workflow problem stemming from tooling limitations, and better visualizations will never close this persistent gap between knowing and doing.
This point is worth lingering over. Humans, by their nature, will adjust to the limitations of the tools they use. But this does not mean that the limitations are ideal, optimal, or even needed. When new tools become available, it opens up new possibilities. This is exactly the situation with the shift from BI to AI. New technology opens up new possibilities for human engagement with that technology. This is the true power of innovation, not merely a new piece of technology, but a new way of doing things that works better for the humans involved at a human level.
Starburst provides access to the data and context to disrupt BI
This gap between knowing and doing remains wide open because the manual effort required to move from a static chart to a real-world decision is a direct symptom of a deeper structural barrier. The friction we experience every day is actually a measurement of distance. The root issue facing modern enterprises is not a lack of data or context. Every organization already possesses more data and information than it can effectively act on.
The true gap is expedient access, specifically the access to the right to that data and context. This lack of time to access inhibits companies across three distinct organizational layers.
Strategic layer
At the strategic layer, a business leader waits days for a critical churn analysis, causing the strategic decision window to close entirely before the insight ever arrives.
Operational layer
At the operational layer, a compliance officer only uncovers critical vulnerabilities during a quarterly audit, preventing the organization from taking immediate corrective action until after an expensive regulatory review.
Infrastructure layer
At the infrastructure layer, a platform manager fails to notice that cluster performance will be severely impacted by increased usage, leaving the technical team unable to respond until users actively report poor performance.
At Starburst, our goal is to eliminate this distance entirely. To realize this vision, we are thrilled to announce that AIDA has officially moved out of public preview and is now Generally Available. It is fully ready for primetime, available today on Starburst Galaxy and via the 480 LTS release for Starburst Enterprise.
How AIDA makes enterprise insights more human
No enterprise hands critical business automation over to an AI system blindly. Trust must be realized progressively, and it serves as the absolute foundation for allowing data to drive a business. To guide this evolution safely, we focus on the AI Trust Ladder, which outlines how an AI assistant progressively earns its place in an enterprise environment.
Stage One: Insights on Demand
Trust begins by delivering accurate, well-reasoned answers in minutes. At this initial stage, the AI provides clear reasoning alongside direct citations of evidence, so users can verify its outputs. AIDA originally launched in April as a public preview built for this exact stage, operating as an elite reasoning agent with federated access and governed answers.
Stage Two: Recommended Actions
Today, AIDA is moving up the ladder to the next phase. In this stage, the AI does not just provide raw data; it actively proposes what to do next, drawing on deep business context. These recommendations must align with your organizational expectations, but human users remain in full control and make the final decision.
Stage Three: Automated Operations
The peak of the ladder focuses on routine workflows with clear corporate guardrails, freeing humans to focus on more complex responsibilities and engage in exceptions only. AIDA launched as an answer engine, but today it has become an enterprise team. If you view AIDA as just another chat agent, you are missing its true potential.
These are not only product development needs – each workflow you want to carry through to automation will go through evaluation of trust at each stage. The rigor and time it takes to climb this ladder of trust will depend on the criticality of the workflow.
What’s coming next for AIDA
AIDA is launching as an operational teammate. Today’s GA release follows six core structural capabilities built for enterprise-scale execution.
Visual narratives with AI dashboards
While text summaries are powerful, sometimes you need visual analytics to help support your next suggested action or decision. AIDA can now generate dynamic AI Dashboards to deliver the clear visual evidence required to justify strategic steps.
Enterprise-grade security guardrails
Deploying AI in production requires absolute safety. Starburst actively protects AIDA against malicious prompt injections and establishes strict operational boundaries to securely scope the AI to its intended enterprise limits.
Composability through custom skills
Organizations can now author their own institutional workflows and step-by-step instructions directly within AIDA. This allows your company to utilize AIDA and MCP for consistent, repeatable execution of complex operational tasks across teams.
Connectivity via third-party MCP Servers
We’re expanding AIDA’s ability to connect to third-party MCP servers. With this, AIDA can now cross the boundaries of the data warehouse to interact with the operational applications your business runs on every day, acting as a client.
AI assurance via feedback and review
Data owners must have complete confidence that the AI is providing accurate, verified results. Because everyday users are the best source of operational validation, AIDA includes built-in feedback loops to capture this critical information and continuously optimize its performance.
Knowledge augmentation with file uploads
To make smarter decisions, agents require specialized context. You can now instantly upload local institutional files, including internal processes, operational run books, and market research, to safely augment the knowledge base of AIDA when you need it most.
How AIDA helps enterprise AI realize real value
When the complete decision workflow is integrated directly into an agentic data layer, exceptional efficiency is unlocked across every core persona in the company.
Proactive regulatory compliance
AIDA is designed to manage workflows that can be created to support regulatory compliance and to support forward-looking activities. For the compliance officer, risk management is no longer a historical, backward-looking exercise.
Resilient supply chains
For the supply chain manager, production lines become entirely resilient to real-world disruptions. When low inventory levels or supply chain anomalies are detected, AIDA automatically triggers component reorders. Manufacturing continues completely unimpeded, significantly minimizing revenue loss.
Robust business infrastructure
For the data engineer, underlying technical issues are identified, and corrective actions are staged well before they cause widespread organizational impact. This proactive posture leads to a radical reduction of costly platform outages.
Empowered knowledge workers
For the average business user, AIDA acts as the dedicated analyst they never had. A complex, multi-day analyst request is compressed into a swift, thirty-second interactive conversation, and the operational action that follows takes minutes rather than weeks. Every worker suddenly gains the analytical depth and operational reach that once required entirely dedicated data teams.
AIDA is centered on the human user
AIDA is designed to make data and insights more accessible, more accurate, more conversational, and ultimately, more human. With AIDA, the timeline collapses dramatically. The human analyst remains fully in charge, asking the core questions and making the strategic calls. AIDA reaches across disparate systems to instantly pull positions from Starburst and relevant unstructured filings directly from SEC EDGAR, while helping the analyst interpret everything that comes back. At this stage, the manual workflow drops to hours, no longer taking days.
What can AIDA become?
We are just getting started on this journey, and we want to share a sneak peek at what is coming to AIDA next. Today, if you unleash generic AI models on federated enterprise data, they frequently hallucinate. This leads to incorrect joins, references to the wrong columns, and incorrect answers.
AIDA is built with a different principle at its core, beginning with its ability to access all the data, governance, trust, and context needed across your entire enterprise. This gives the governed context to prevent hallucinations. Closing that gap completely is our core work, allowing the user to bring the question while the system seamlessly handles the rest.
AIDA is built to maximize accuracy and trust
Because trust must be earned rather than assumed, we are introducing Context-Aware Semantic Layer Integration and User-Defined Evals. These features will ensure AIDA integrates your unique business context to improve accuracy against the specific benchmarks you define.
Enterprise AI governance
Enterprise AI is fundamentally different from consumer AI. Maintaining absolute consistency, managing AI components and assets, and keeping an immutable record of every action taken is critical for your organizational success. We will deliver these rigorous controls through centralized Skills Management and Enhanced Guardrails.
Welcoming AIDA to the team
AIDA is designed to make enterprise AI function in practice. In doing so, it solves one of the core challenges of BI workloads by disrupting the static, slow-moving nature of dashboards and replacing it with an interactive, dynamic, conversational approach.
This approach is more than just a new interface. It’s also a human-centric interactive model. In essence, AIDA is a new member of the team, and just like a new teammate, its addition to your organization provides new opportunities for expansion and growth. This is exactly what we’re seeing in action with AIDA as it joins enterprise environments.
Try AIDA today
We would love AIDA’s next organization to be yours. You can turn on AIDA in Starburst Galaxy today, update to the 480 LTS release in Starburst Enterprise, or connect with our team to get started.
