What does Ronja actually do?

Ronja is a governed, last-mile execution layer for self-serve AI analytics. It connects to your data warehouse, data lake, and 100+ SaaS apps, brings all of it under one governed layer, and lets anyone in the organization get traceable answers through Slack, ChatGPT, Claude, or Ronja directly. Every number traces to its source: real code, real rows, real lineage. No hallucinations.

How is Ronja different from a traditional BI tool?

Traditional BI tools like Tableau or Power BI require analysts to build dashboards that business users then consume. Ronja removes the middle step: business users ask questions in plain language and get governed, traceable answers instantly. Ronja layers on top of your existing stack: the warehouse stays, dbt stays, existing dashboards stay. It handles the routine questions that currently sit in the data team's backlog.

Does Ronja replace our data team?

No. Ronja is the data team's first line of support. It handles routine ad-hoc questions so the data team can focus on the hard problems that require human judgment. Every interaction surfaces what the business actually needs: data gaps, divergent definitions, and common patterns. The data team goes from reactive firefighting to proactive, strategic work.

What about warehouse costs when everyone starts querying data?

Ad-hoc queries run on Ronja's own execution layer (DuckDB on Parquet/S3). They never hit the warehouse. Adding more users does not increase your warehouse bill. Your warehouse continues to handle what it was designed for: heavy transformations, scheduled pipelines, and system-of-record storage.

How does Ronja ensure answers are accurate and consistent?

Ronja maintains a federated context layer that aggregates definitions from wherever they already live: dbt models, Cube, Confluence, warehouse metadata, and user conversations. It detects where definitions diverge, learns from every correction, and ensures the same question always gets the same answer, regardless of who asks or which channel they use. Every answer traces to source with full lineage.

Self-serve analytics that actually works. Governed, traceable, and built to scale.

Ronja is a governed, last-mile execution layer that sits between your existing data infrastructure and the people who need answers. Fully traceable AI analytics. No hallucinations.

Get started See how it works

Ronja layers on top of your existing data stack. Your data team sets the rules. Everyone else gets instant, governed answers.

Ronja speaks your language. Ask a question the way you'd ask a colleague. Agents translate it into real, auditable analysis behind the scenes.

Endorsed data sets let anyone propose a definition and authorized users approve it. Once endorsed, agents use that definition everywhere. No more conflicting numbers.

Ronja's agents build dashboards and analyses using endorsed definitions. From question to governed result in minutes, no ticket required.

Every answer comes with full data lineage. Click any number to see exactly where it came from, how it was calculated, and when it was last updated.

Ronja plain language chat analysis interface

Ronja endorsed metric definition for Production Loss Cost

Ronja CFO dashboard with KPIs and revenue trends

Ronja handles routine ad-hoc questions and surfaces what the business actually needs: data gaps, divergent definitions, and common patterns. Your team gets a demand-driven roadmap and focuses on strategic work.

Your team shifts from executing ad-hoc requests to governing data quality, endorsing definitions, and setting the rules agents follow.

Agents monitor pipeline health 24/7 and rebuild when upstream data changes. You review and approve.

Ronja eliminates the translation step. Business users describe what they need in plain language, and agents generate real, versioned, auditable code.

Ronja home overview with databases and explorations

Ronja Finetune governance and review interface

Ronja pipeline flow view with agent maintenance

Ronja auto-generated table transformation logic

The intelligence flywheel

The more you use it, the smarter it gets

Ronja layers on top of your existing stack and builds institutional knowledge over time.

Define what a metric means and agents build the data behind it.

Connect a new source and Ronja documents and understands it automatically.

Correct a number and Ronja traces the impact across every table and dashboard.

Review what changed — approve, reject, or refine before anything goes live.

It learns your language

What's our MRR?

I don't have a definition for MRR yet. How does your team calculate it?

9:41 AM

Sum of all active subscriptions, excluding free trials and annual prepayments divided by 12

Definition saved

MRR = Sum of active subscriptions, excl. free trials, annual ÷ 12

monthly_recurring_revenue

2 days later

What's our MRR?

Your MRR is €142,800 as of March 1.

Using endorsed definition

9:43 AM

It watches your data

4 improvement opportunities

HubSpot sync stalled 2 min ago

Last successful sync was 6 hours ago. 340 contacts may be outdated.

New Stripe fields documented 1 hour ago

12 new fields detected in the invoices endpoint. Descriptions auto-generated and pending review.

Missing CAC data for Q4 3 hours ago

Customer acquisition cost can't be calculated — marketing spend source not connected.

Duplicate customer IDs detected 5 hours ago

14 records in the customers table share the same external ID from Salesforce.

It adapts to corrections

Actually, MRR should exclude customers in grace period too

Got it. I've updated the MRR definition and run a quality check.

↻ Definition updated

− Sum of active subscriptions, excl. free trials, annual ÷ 12

+ Sum of active subscriptions, excl. free trials, grace period, annual ÷ 12

Updated by @anton · Applied to 3 tables

⚠ Quality check triggered

Found 23 customers currently in grace period. 8 were included in last month’s MRR calculation.

Recalculated MRR: €138,200 (was €142,800)

Your team stays in control

Metric definition Requested by @lisa

Update MRR to exclude grace period customers

Lisa from Finance suggested this change after Q4 close. Currently 23 customers in grace period are included.

What will change

Definition monthly_recurring_revenue

Affects 3 tables

Dashboards Revenue Overview Board Report

Impact −€4,600/mo

Data source Requested by @erik

Add Salesforce as primary CRM source

Erik from Sales wants to connect Salesforce to replace the manual CSV imports for pipeline data.

"The more we use Ronja, the better it understands our organization and my intents."

Once a definition is endorsed, it applies everywhere. Same question, same answer. And every question asked is a signal: Ronja surfaces data gaps, divergent definitions, and the transformations your business uses most. Your data team gets a roadmap written by real usage.

The shift

The future of self-serve analytics is agentic, governed, and accurate. Data stacks were built for specialists. Ronja adds the last mile: a governed execution layer where agents handle routine data requests, every answer traces to source, and the data team finally gets to do strategic work instead of running a help desk. And every question asked is a signal, surfacing what the business actually needs, so the data team knows exactly what to build next.

AI-Native

See what your data can do when it's governed, traceable, and instantly accessible

From question to traceable answer. Without a ticket, without a dashboard, without hitting the warehouse.

cloud.ronja.tech/datastudio

cloud.ronja.tech/dashboard

How Ronja compares

Built for agents, and humans

	Ronja	Data PlatformsSnowflake, Databricks	BI ToolsPower BI, Tableau	AI Query ToolsHex, Julius
Own execution layer (no warehouse compute)	Yes	No	No	No
Layers on existing stack	Yes	N/A	Needs warehouse	Needs warehouse
Who does the work	Agents + data team (agents handle routine, team handles strategy)	Your data team	Your analysts	You ask, AI queries
Accuracy	Deterministic, traceable	Depends on your team	Depends on the model	May hallucinate
Time to insight	Minutes	Weeks to months	Days to weeks	Seconds, but fragile
Data team required	Yes. Fewer tickets, more strategic work	Yes	Effectively yes	No, but limited

Security & governance

Secure by design. Traceable by default.

Trust starts with transparency. Ronja is built on real code, real governance, and real transparency. Every number traces back to its source. Agents write real, auditable code. No black boxes. Ever.

Access control

Control who sees data, code, and outputs. At every level.

Data lineage

Every metric traces back to its source. Fully auditable.

Compliance

Enterprise-grade security standards.

No black boxes

Agents write real code. Auditable, versioned, reproducible.

Self-serve analytics that actually works. Governed, traceable, and built to scale.

Ask in plain language, get real answers

One definition for every metric

Dashboards built in minutes, not weeks

Every number traceable to source

Your team's first line of support

Govern the logic, not the queue

Monitor, rebuild, repeat. Automatically.