Tasks

Tasks are interactive execution environments built into the platform. Each task gives you a persistent Python or TypeScript sandbox with your catalogs, lake tables, and external connections already wired in - no credentials to configure, no installs to run.

Python
TypeScript

The Python sandbox runs Python 3.13 and comes with the following libraries pre-installed:

Library	Purpose
`polars`	DataFrame processing
`pandas`	DataFrame processing
`duckdb`	In-process SQL
`pyiceberg`	Iceberg catalog client
`numpy`	Numerical computing
`matplotlib`	Plotting
`pyarrow`	Arrow/Parquet I/O

The `oleander` module

Every cell has access to a generated oleander module that wires your catalog and DuckDB connection automatically:

# Pre-wired DuckDB connection with all your Iceberg catalogs attached
conn = oleander.conn

# Default Lakekeeper catalog as a pyiceberg catalog object
catalog = oleander.default_catalog

# Get any registered catalog by name
my_catalog = oleander.get_catalog("my_catalog_name")

Query your lake directly:

result = conn.execute("SELECT * FROM oleander.default.flowers LIMIT 10").df()
result

Use pyiceberg for low-level catalog operations:

table = catalog.load_table("default.flowers")
df = table.scan().to_pandas()
df.head()

BigQuery

If you have BigQuery connections configured, the DuckDB bigquery extension is installed automatically. BigQuery tables are accessible as connection_name.dataset.table:

result = conn.execute("""
  SELECT date, SUM(revenue) AS total
  FROM my_bq.analytics.orders
  WHERE date >= '2024-01-01'
  GROUP BY 1
  ORDER BY 1
""").df()

State persistence

Variables persist across cell executions within a session. If you define a DataFrame in one cell, it is available in the next. State is stored in /tmp/_oleander_state.pkl and survives cell reruns.

Output

DataFrames are displayed as structured tables. Matplotlib figures are rendered as inline SVG. Other values are printed as text.

The TypeScript sandbox runs Node 24 with tsx for execution. A generated oleander module exposes your Iceberg catalogs:

import { defaultCatalog, getCatalog, catalogs } from './oleander';

// List all catalogs
console.log(catalogs.map(c => c.name));

// Load a table from the default catalog
const table = await defaultCatalog.loadTable({ namespace: ['default'], name: 'flowers' });

Code generation

Each cell has an assist button that generates code based on your prompt, aware of your available libraries and catalog structure. Generated code appears as a new cell ready to run.

Spark cells

Submit inline PySpark code as a managed Spark job directly from a task cell. The job runs in the oleander.tasks namespace and returns a run ID you can track in the platform. Lineage from Spark cell runs is captured automatically.

Getting Started

Platform

Coding with agents

Connections

CLI

The `oleander` module

BigQuery

State persistence

Output

Code generation

Spark cells

​The oleander module

​BigQuery

​State persistence

​Output

​Code generation

​Spark cells

The `oleander` module

BigQuery

State persistence

Output

Code generation

Spark cells