Skip to content

Welcome to Kyomi: Bridging Chat and Dashboards

November 16, 2025

We're excited to introduce Kyomi, an AI-powered analytics platform that solves a problem we kept running into: why do we have to choose between simple chat interfaces and powerful dashboards?

The Analytics Dilemma

If you've worked with data recently, you've probably noticed a divide:

On one side: AI chat tools like ChatGPT's data analysis or simple analytics chatbots. They're incredible for quick questions - just type "show me revenue by region" and get an instant answer. But the moment you want to save that insight, customize it, or share it with your team? You're stuck. Chat interfaces don't scale beyond the conversation.

On the other side: Traditional BI tools like Tableau, Looker, or Metabase. They're powerful, customizable, and perfect for production dashboards. But they require upfront configuration, schema mapping, and often weeks of setup before anyone can ask a simple question. The barrier to entry is high, and iteration is slow.

The result? Teams end up with two separate tools:

  • Chat for quick ad-hoc questions (throw away the results)
  • BI for dashboards (takes weeks to build)

But what about the space in between? What happens when:

  • You ask a question in chat and want to save it as a dashboard?
  • You build a dashboard but want to ask follow-up questions conversationally?
  • You start simple but need to customize as requirements grow?

Enter Kyomi

Kyomi bridges this gap. It's designed around a simple principle: your analytics workflow should be seamless, from question to dashboard.

Here's how it works:

1. Start with a Question

Type your question in plain English. No SQL required, no schema to learn. Just ask:

"Show me our top 10 customers by revenue this quarter"

2. AI Finds the Answer

Kyomi's AI agent:

  • Finds the right tables in your data warehouse (BigQuery)
  • Writes optimized SQL
  • Generates visualizations automatically
  • Explains the results

3. Refine Conversationally

Don't like the chart? Want to filter by region? Just ask:

"Can you break that down by region and show it as a stacked bar chart?"

The AI updates everything in real-time.

4. Save to Dashboard

Found something valuable? One click saves it to a dashboard. No rebuilding in another tool, no exporting and importing. The transition from exploration to production is instant.

5. Customize When Ready

Need more control? Every dashboard is built with ChartML, our open-source dashboard language. Edit the code directly, add custom layouts, create interactive filters - full control when you need it, simplicity when you don't.

6. Share with Your Team

Publish dashboards, add filters, collaborate. Everyone works from the same live data, with the option to ask follow-up questions conversationally.

What Makes Kyomi Different?

It's not just chat. You get real dashboards that you can customize, share, and iterate on.

It's not just BI. You can start with a conversation, no setup required.

It learns your business. Define metrics once (like "Active Customers" or "MRR"), and Kyomi applies your definitions consistently across all queries and dashboards.

No vendor lock-in. Dashboards are markdown files with ChartML. Your data stays in your warehouse. You can export everything.

Free tier that works. Full SQL editor and up to 5 dashboards, free forever. Only pay for AI features when you're ready.

The Vision

We believe analytics shouldn't force you to choose between speed and power. The best tools adapt to how you work:

  • For quick questions: Chat interface, instant answers
  • For exploration: Conversational refinement, visual feedback
  • For production: Code-based dashboards, full customization
  • For teams: Sharing, collaboration, consistent definitions

All in one seamless workflow.

Try It Yourself

Kyomi is live and ready to use. Connect your BigQuery warehouse and start asking questions in minutes.

Ready to bridge the gap?

Start free with trial AI features. No credit card required.

Get Started Free →

What's next? In upcoming posts, we'll dive deeper into:

  • How the AI agent works under the hood
  • Building custom dashboards with ChartML
  • Best practices for defining workspace knowledge
  • Real-world use cases from early users

Have questions or feedback? We'd love to hear from you. Reach out at hello@kyomi.ai.


Back to Blog