From your data sources to a verified answer in chat.
oneAgent connects your existing systems, maps business terms and metrics to your data logic, and makes verified answers usable via chat, API, or dashboard — live against the source system or through the oneAgent Lake.
Test environment with prepared shop data · no credit card · no data connection required
Net revenue per customer in 2024?
| Müller GmbH | €128,400 |
| Schmidt KG | €96,700 |
| Becker AG | €81,250 |
Data sources → oneAgent Control Layer → verified chat answer
oneAgent connects data, metrics, and AI in one controlled workflow.
AI is not where company figures are freely invented. It helps understand questions and make results usable. The calculation follows your data sources, metrics, filters, time references, and business rules.
Connect data sources
ERP, CRM, shop, data warehouse, databases, files, or APIs — live or via import into the oneAgent Lake.
Map business logic
Metrics, aliases, relationships, filters, time references, and business rules are set up together with the business team, controlling, BI, or data team.
Verify & approve metrics
Key metrics are checked with test cases, comparison against the source system, and validation, then approved as trusted.
Ask questions & use results
Users ask questions in natural language and get tables, charts, summaries, ideas, forecasts, or dashboard building blocks.
Business teams work in natural language. The data logic stays controlled.
See why oneAgent →Connect the data sources your company already uses.
oneAgent can make ERP, CRM, shop, data warehouse, database, file, and API data usable. Your existing systems stay authoritative — oneAgent makes them accessible via chat and workflow.
ERP and inventory management
Order, revenue, inventory, purchasing, and financial data from existing systems.
CRM and sales
Customers, leads, opportunities, activities, segments, and pipeline metrics.
Shops and e-commerce
Orders, products, categories, returns, customers, inventory, and campaign references.
Data warehouse & databases
Existing models, tables, views, and structured data from your data platform.
Files and APIs
CSV, Excel, or API data when operational or external sources need to be added.
Example connectors
Two ways to your data: live or in the oneAgent Lake
Depending on your data landscape, freshness needs, and performance goals, oneAgent queries your systems directly — or brings selected data into the oneAgent Lake via import / ETL.
Live connection
oneAgent connects directly to your data warehouse, database, or a source system. Data doesn't need to be copied. The query runs live against the existing infrastructure.
Well suited for
Import / ETL
oneAgent takes selected data from source systems, transforms it, and makes it available in the oneAgent Lake for fast analysis. Useful when data from multiple sources needs to be combined, prepared, or queried at scale.
Well suited for
oneAgent translates your language into your data logic.
Business teams don't ask for table and column names. They ask about revenue, margin, return rate, customers, products, regions, or time periods. The semantic layer maps these terms to the right data objects, metrics, and relationships.
Recognize terms and aliases
"Revenue", "turnover", or "monthly revenue" can all map to the same approved metric.
Map entities
Customers, products, orders, countries, channels, or campaigns are linked to the right tables, columns, and relationships.
Understand relationships
oneAgent knows how orders, customers, products, time periods, or campaigns relate, and which join or filter logic applies.
Consider context
Time periods, filters, role permissions, and business rules all flow into the analysis.
Defined once. Safely reused in chat.
Metrics like revenue, margin, return rate, inventory, or contribution margin aren't freely generated by AI. They're set up as controlled logic: with source, calculation, aliases, owners, time reference, required filters, and a description. oneAgent then recognizes the metric in natural-language questions and uses it reproducibly in chat.
What was net revenue in the online shop in 2024?
€4.82M
Net revenue · Online shop · Fiscal year 2024
In chat, oneAgent doesn't just recognize a word like "revenue" — it uses the previously defined metric, including source, time reference, and filters.
Business teams ask in natural language — oneAgent delivers usable results.
Users don't need to know the table structure or write SQL. They ask questions like in chat and get results based on the configured metrics, data sources, and rules.
See use cases by team →
New chatoneAgent·Yesterday·12:43 PMResult formats
Table
Detailed analyses, rankings, customer segments, product lists, or deviations.
Chart
Trends, comparisons, developments, and management overviews.
Summary
Quick management briefings or explanatory context.
Ideas & next steps
Campaign ideas, action recommendations, or root-cause hypotheses.
Forecast
ML-based forecasts based on historical data — forecasts, not verified actuals.
Dashboard building block
Recurring questions that should be saved and reused.
Self-service for business teams. Control for IT, BI, and data teams.
oneAgent is designed for company data. Access, roles, data sources, metrics, and business rules can be set up in a controlled way. Depending on requirements, operation in the cloud, on your own infrastructure, or on-premise is possible.
No black box. No magic promise. A controlled data and AI layer.
oneAgent aims to make self-service easier — but should never suggest that data quality, metric logic, role permissions, and validation happen automatically without business ownership.
Not: AI magic
Instead: a controlled process
Frequently asked questions
Answers about data connections, metric logic, and verified results.
No. oneAgent can complement existing databases, data warehouses, and source systems. Depending on the use case, it connects live or imports a selected data set into the oneAgent Lake.
No. Business teams ask questions in natural language. The technical data logic, metrics, filters, and relationships are set up in a controlled way in the background.
oneAgent uses AI to understand the question. The calculation, however, follows defined metrics, data sources, filters, time references, and business rules. Key metrics can be verified and approved as trusted.
Yes. The free trial runs for 14 days with prepared shop data, predefined metrics, and guided use cases. Your own data is currently connected together as part of a supported pilot project.
See how data sources become verified answers.
Try oneAgent for free with prepared shop data — or book a demo directly if you'd like to discuss data sources, connection types, security, or a pilot with your own data.
For pilot projects, we set up an initial data source, verified metrics, and concrete use cases together with you. Connect your own data
Not sure what fits? Compare trial, demo & pilot