Query
Just select the table
and the conditions
Select the schema, table and columns, then add WHERE conditions.
Features
Register data sources, connect Tasks on a canvas, and run.
01
Data sources
Register them under Tools ▸ Data Sources.
| Type | Also install | What you enter |
|---|---|---|
| PostgreSQL | nothing | host, port, database, user, password, TLS mode |
| MySQL / MariaDB | nothing | host, port, database, user, password, TLS mode |
| SQL Server | nothing | host, port, database, user, password, TLS mode |
| Oracle | nothing | host, port, service name, user, password |
| Delimited file | nothing | the file (csv, txt, tsv, dat), delimiter, encoding, first data row |
- DBMS
- PostgreSQL
- MySQL
- MariaDB
- SQL Server
- Oracle
- Db2
- Snowflake
- BigQuery
- Databricks
- Redshift
- ClickHouse
- Trino
- CockroachDB
- TiDB
- SingleStore
- SAP HANA
- Teradata
- MongoDB
- SQLite
- DuckDB
- Graph databases
- Neo4j
- Files
- CSV
- TXT
- TSV
- DAT
- Excel
- API
- Use data provided by an API as a data source.
Just select the table
02
Canvas, Tasks, SubTasks
Place Tasks and connect them with lines to complete a workflow. SubTasks process a Task's result step by step.
| Task | Description |
|---|---|
| Query | Queries a table of a data source, with conditions (WHERE). |
| File | Reads a delimited file or pasted text. |
| Join | Joins two results on the columns you specify. |
| Merge | Aligns the columns of two results and combines them. |
| Difference | Extracts the rows found on one side only. |
| Query Drive | Uses the result values of the previous Task as conditions of the next Query. |
| Table | Uses the result of a completed run as an input. |
| UDF Query | Reuses a saved canvas as a Task. |
| Report | Outputs the result. |
| SubTask | Description |
|---|---|
| Filter | Extracts only the rows that match a condition. |
| Sort | Sorts by the columns you specify. |
| Summary | Groups by key columns and calculates sum, count, maximum and minimum. |
| Calculate | Creates new columns from expressions. |
| Output Columns | Sets the output columns, their order and aliases. |
| Pivot | Converts row values into columns. |
SubTask
Apply filter, summary and sort
in order inside one Task
The output columns of each step are shown before you run.
- Filter
- Sort
- Summary
- Calculate
- Output Columns
- Pivot
03
Runs and schedules
Workflows run on your PC.
- ValidateAlt + F5
- Checks for missing items and errors before a run.
- RunF5
- Runs the whole workflow.
- Partial RunShift + F5
- Runs only the selected Report and the Tasks connected to it.
- ScheduleScheduled run
- Set the interval and the time zone, and the schedule runs even after Turbrix is closed.
- Run historyExecution Explorer
- Records the start and end time and the status of every run; earlier results can be opened again.
04
Results and output
View results as grids and charts, and save them as files.
Report
See the result at once
as a grid and charts
Save it as CSV, SQLite or Parquet files, or open it in Excel.
| Output | Description | Edition |
|---|---|---|
| Result preview | Shows the first 200 rows as a grid. | free · paid |
| Browse full result | Supports paging, sorting and condition filters. | free · paid |
| CSV | Saves the whole result as a file. | free · paid |
| SQLite, GZIP | Created together with the CSV when selected in the output settings. | free · paid |
| Charts and reports | Saved as an HTML report, PDF, PNG or JPG. | free · paid |
| Excel | Opens the result CSV in the Excel installed on the PC. | free · paid |
| Parquet | Saves a file that keeps the column types. | paid |
Parquet
Save as Parquet files
Set compression, file splitting and partition folders.
Preparing training data
Apply train / validation / test splits, sampling, missing-value handling, encoding and normalisation as SubTasks. The values used for each transformation are saved as a file.
LLM datasets
Map columns to roles and save them as JSONL.
- messages JSONL
- Alpaca
- ShareGPT
- Hugging Face
- RAG chunks
PII masking
Select hashing, tokenisation or partial masking for each column.
Power BI
Open the result in Power BI Desktop.
05
Storage
Saved canvases and results are kept in the data folder on your PC, %LOCALAPPDATA%\Turbrix\Local\data.
Options ▸ Storage
Set how long results are kept
06
AI agents
AI agents can do what you do on the screen.
MCP
Build and run workflows in a conversation
Turbrix provides its features as MCP tools, so AI agents can work with it. The in-app chat window runs on your own AI API key.
- create a workflow
- edit
- run
- recommend
- explain
- summarise a result
- AI agent
- MCP tools
Turbrix canvas
07
Team collaboration
A team works together by connecting to a Turbrix installed on a company server.
- Server mode
- Select server mode in the options to work with the data sources registered on the server.
- Accounts and roles
- An administrator manages data sources and metadata; users sign in with company accounts (SSO).
- Cloud
- Runs on a dedicated server.
- Enterprise
- Installed and run inside your closed network.
Contact: turbrix@seizia.com
08
Windows and macOS
- Windows
- Windows 10 · 11 (64-bit)
- macOS
- Apple silicon, macOS 13 or later