What’s new in Zebra AI: smarter chat, financial data support, and Databricks & Snowflake

Zebra AI is evolving fast and this update brings some of the most-requested improvements yet. From smarter chat reasoning and first steps in automated financial reporting, to expanded database support with Databricks and Snowflake, these updates make it easier than ever to turn your data into decisions.

Smarter, deeper chat 🎯

Until now, Zebra AI chat was built on a fixed, three-step process: you asked a question, the system followed a predefined flow, and it gave you an answer. Now it can do so much more!

With this update, we’ve rebuilt the chat around a single, more powerful agent that can reason iteratively. Instead of stopping at the first answer, it thinks, tests, adjusts, and goes deeper until it finds a meaningful result.

What’s new in practice:

• One flexible agent instead of three fixed ones – it adapts dynamically instead of following a rigid flow
• Iterative reasoning – the agent breaks down complex questions into steps, evaluates the results, and keeps going until it reaches a reliable conclusion
• Self-correction – if it runs into an error, it can fix it and try again
• Code interpreter built in – the agent can generate new calculations, transform datasets, or create entirely new “data slices” to answer your questions

What does this enable?

• Ask “Why did the women’s category underperform?” → Zebra AI doesn’t just stop at one explanation. It explores multiple dimensions: price, volume, timing, regional breakdowns, even manager performance and only then delivers a complete answer.
• Upload a dataset with sales and margins → Zebra AI can calculate missing metrics like costs, then create visuals automatically.
• Have data until August → Zebra AI can generate forecasts for the rest of the year, and even propose optimistic vs. pessimistic scenarios.
• Looking for deeper insights → it can perform advanced analyses like price-volume-mix or ABC analysis without needing manual setup.

Why it matters?

✅ More accurate and reliable answers, not just quick guesses

✅ Goes far beyond your raw dataset from derived metrics to forecasts

✅ Handles multi-step, complex, “why” and “what if” questions that normally take analysts hours

✅ Keeps context across your session, so you can build on previous results (“Now increase the forecast by 5% and show me the effect on gross margin”)

With this upgrade, Zebra AI isn’t just answering questions, but it’s reasoning, calculating, and predicting. That means you can move from descriptive analytics (what happened) to diagnostic (why it happened) and predictive (what could happen next) in a single conversation.

A note on limitations:
Because the new agent carries out a deeper analysis, some answers may take a bit longer. And like all AI systems, it can occasionally misinterpret or misaggregate data. The good news? It usually corrects itself if you ask again or point out the mistake.

👉 Try it out on your own data and see how much deeper you can go.

Automated P&L reporting as the first step toward financial workflows

Financial reporting has been the number one request from Zebra AI users and it’s now supported in its first version.

When you upload a P&L dataset (Excel, Power BI, SQL, etc.), Zebra AI recognizes it and builds a structured report automatically:

• Detects income statement structures
• Classifies accounts and highlights key rows (e.g. Revenue, COGS, Net Income)
• Applies Zebra BI Tables with built-in waterfall logic
• Adds trend visuals for quick review (e.g. revenue vs. prior year)

Why this matters:

✅ Skip manual setup and get a clear P&L in seconds

✅ Share reports immediately with colleagues

✅ Works for both SMB-style Excel exports and more granular enterprise datasets

Try it on your own P&L export, and if you’d like to explore deeper financial workflows, connect with us at support@zebra-ai.com.

Zebra AI + Databricks & Snowflake = conversational analytics at scale

Zebra AI now connects directly to Databricks and Snowflake, joining our existing support for Power BI, SQL Server, MySQL, and PostgreSQL.

For BI & Data Teams: 

• Offer true self-service analytics without building or maintaining dashboards 

• Stay in control, define the right tables, filters, and access 

• Reduce ad-hoc requests and free up time for strategic work 

For Business Decision-Makers: 

• Ask questions in natural language, get instant answers and visuals 

• Explore different scenarios and validate assumptions on the fly 

• Reduce time-to-insight from days to seconds 

Databricks and Snowflake are central data platforms for many enterprises. With Zebra AI, business users can finally explore them conversationally, while BI teams stay firmly in control.

Want to see how Zebra AI works with your Databricks environment? Reach out at support@zebra-ai.com to explore a proof of concept.

Questions? Feedback? We’d love to hear from you, so write your message to support@zebra-ai.com.