Dashboards vs AI Agents: Why the Future of Analytics Needs Both

Editorial note: This blog is based on a recent podcast between Joe Reis and Michael Driscoll on the evolving landscape of business intelligence and data engineering. 

Every few years, the analytics industry declares the death of dashboards.

First it was self-service BI.
Then embedded analytics.
Now it’s AI agents.

With the rise of large language models and conversational interfaces, a new narrative has emerged:

“Dashboards are obsolete. AI will answer all your data questions.”

It’s a compelling idea.

Instead of navigating complex dashboards, users could simply ask:

• Why did revenue drop yesterday?
  
  
• What’s driving the spike in signups?
  
  
• Which customer segments are growing fastest?
  

AI generates the query, analyzes the results, and returns an explanation.

No dashboards required.

But there’s a problem with this vision.

Analytics isn’t just about getting answers.

It’s about trusting them.

And that’s where dashboards remain essential.

The Hidden Role Dashboards Play in Analytics

Dashboards are often dismissed as static reporting tools.

But their real function is deeper: they create shared context.

A dashboard defines:

• what a metric means
  
  
• how it’s calculated
  
  
• which data sources are used
  
  
• how time is aggregated
  

When a team looks at a dashboard together, everyone sees the same truth.

Without that shared context, analytics quickly becomes chaotic.

One team’s “revenue” might exclude refunds.
Another might include taxes.
A third might calculate it differently altogether.

Dashboards—and the semantic layers behind them—solve this problem.

They establish a common language for metrics.

That role becomes even more important in an AI-driven world.

AI Agents Change How We Ask Questions

Where dashboards provide structure, AI agents provide speed.

Instead of navigating dozens of charts, users can simply ask questions.

AI systems translate natural language into queries, run analyses, and summarize results.

This changes analytics in two major ways.

1. Analytics becomes accessible to everyone

Historically, analytics required specialized skills.

Users needed to understand dashboards, SQL, or data models.

AI removes much of that friction.

Anyone in an organization can now ask questions about data.

2. Discovery becomes dramatically faster

Instead of navigating multiple dashboards to investigate a problem, users can jump directly to explanations.

AI can surface correlations, anomalies, and trends in seconds.

This is why conversational analytics is gaining momentum across the industry.

But speed alone isn’t enough.

The Trust Problem in AI Analytics

AI systems can generate answers quickly.

But users still need to verify those answers.

If an AI agent says:

“Revenue declined due to lower conversion rates in Europe.”

Most people won’t immediately act on that insight.

They’ll want to see:

• the revenue trend
  
  
• the regional breakdown
  
  
• the conversion rate change
  
  
• the underlying data
  

In other words, they want visual evidence.

This is where dashboards and visual analytics become essential.

Charts provide the context needed to confirm—or challenge—AI-generated conclusions.

Without that transparency, analytics risks becoming a black box.

Why the Semantic Layer Is Making a Comeback

Another trend resurfacing alongside AI is the semantic layer.

For AI to analyze data reliably, it needs structured definitions of metrics.

Otherwise, it must interpret raw tables and guess how metrics should be calculated.

That’s risky.

A semantic layer solves this by defining:

• business metrics
  
  
• data relationships
  
  
• aggregation rules
  
  
• metric definitions
  

This ensures that AI queries and dashboard queries produce the same results.

In other words, the semantic layer becomes the foundation for trustworthy AI analytics.

The Real Bottleneck: Data Engineering

Despite the excitement around AI interfaces, the hardest problems in analytics remain the same.

Data must still be:

• ingested
  
  
• cleaned
  
  
• transformed
  
  
• modeled
  
  
• made queryable
  

If the underlying data pipeline is unreliable, every analytics interface fails.

Dashboards break.
AI answers become inconsistent.
Metrics lose credibility.

This is why the most important layer in the modern analytics stack is still data engineering.

AI can improve how we explore data—but it cannot fix poorly structured systems.

The Future Analytics Workflow

Instead of replacing dashboards, AI agents will reshape how we interact with them.

A typical analytics workflow might look like this:

1. Ask a question via conversational AI
   
   
2. Generate an explanation automatically
   
   
3. Validate the insight using visual dashboards
   
   
4. Investigate deeper through segmentation and filtering
   
   
5. Automate monitoring using AI agents
   

This loop combines the strengths of both interfaces.

AI accelerates discovery.

Dashboards provide clarity.

Together, they create a far more powerful analytics experience.

The Real Goal of Analytics

Analytics tools often focus on interfaces—dashboards, notebooks, or AI agents.

But the real goal is much simpler:

Help people understand what’s happening in their business fast enough to act.

AI agents will make analytics dramatically easier to access.

Dashboards will keep insights grounded in evidence.

The future of analytics isn’t about choosing between them.

It’s about building systems where humans and AI explore data together.

FAQ

Will AI replace dashboards?

No. AI agents will complement dashboards by accelerating data exploration, while dashboards provide visual validation and shared context for teams.

What is conversational analytics?

Conversational analytics allows users to ask questions about their data in natural language. AI translates those questions into queries and returns insights automatically.

Why are dashboards still important?

Dashboards create shared definitions for metrics and allow users to visually validate insights, which helps maintain trust in analytics systems.

What is the semantic layer in analytics?

The semantic layer defines business metrics, relationships between datasets, and aggregation rules, ensuring consistency across dashboards, queries, and AI tools.