unwind

A DuckDB-only SQL/Python orchestrator with Jinja templating, a semantic layer, and lineage at the table, column, and value level. An alternative to Dbt and SQLMesh, focused on deterministic value-lineage backed by a SQLGlot AST and DuckDB as the single execution engine.

A project is a directory of .sql and .py files. SQL models give full table/column/value lineage via the SQLGlot AST. Python models are opaque to lineage but participate in the DAG like any other node — typical use cases are sources (you bring your own Postgres/Oracle/S3/CSV ingestion code) and sinks (exporting a final table elsewhere). The two kinds compose freely.

Unwind deliberately ships zero DB-third-party dependencies. If you want to pull from Postgres, you pip install psycopg yourself; if you want Oracle, you pip install oracledb. Unwind's only DB is DuckDB.

Status: alpha — table, column, and deterministic value lineage all work end-to-end, plus a web UI and an optional LLM investigator (pydantic-ai, multi-provider).

Install

The core install (duckdb, jinja2, sqlglot) is enough to load a project, plan its DAG, run it on DuckDB, and compute table / column / value lineage in Python. The two surfaces that reach beyond the core — the web UI and the LLM investigator — live behind optional extras.

Pick what you need:

pip install unwind-sql                    # core only
pip install "unwind-sql[web]"             # add the web UI
pip install "unwind-sql[web,llm]"         # web UI + LLM investigator
pip install "unwind-sql[all]"             # everything — recommended for trying it out

For local development on this repo:

uv sync    # core + the dev group (which includes web/llm bits and pyarrow)

Run the example

The bundled example computes net margin per order over 5 raw/ref tables, with a Jinja macro and a final aggregation by warehouse:

cd example/
uv run python generate_data.py   # one-time: write the 5 parquet sources
uv run python main.py

The script:

1. Loads models/ (mixed .sql + .py) and runs the full DAG on DuckDB.
2. Prints the table & column lineage of int_tax_costs.local_tax_fee.
3. Traces local_tax_fee for order_id="ORD-7892" back to the raw values that contributed (raw_orders.gross_sales = 500.0, ref_local_taxes.tax_pct = 0.20, …).
4. If OPENAI_API_KEY is set, asks an LLM to explain the trace in plain language and flag suspicious values (uses pydantic-ai, swap providers by passing llm_provider="anthropic" etc.).
5. Opens a browser tab on the web UI (React Flow / @xyflow/react DAG
   • per-column lineage tree, dagre layout). Press Ctrl+C to stop.

raw_orders in the example is a Python model (example/models/raw_orders.py) that reads the bundled parquet fixture via pyarrow. To wire it to your own source (Postgres, Oracle, S3, REST API, …) edit example/models/helpers.py — Unwind has no built-in connectors, so you import the lib you want and call it yourself.

See example/main.py and example/README.md for the model walkthrough.

Python models

A file in models/ whose Python module defines a top-level callable model(context) is recognised as a node of the DAG. Anything else in models/*.py is imported as a plain helper module — so from helpers import load_data just works.

# models/raw_orders.py — Arrow-native ingestion, zero-copy into DuckDB
import pyarrow.parquet as pq

GROUP = "costs"
MATERIALIZED = "view"     # "table" (default) or "view"
DEPENDS_ON = ()           # tuple of upstream model names

def model(context):
    # context.connection (live DuckDBPyConnection),
    # context.variables (Jinja vars), context.project_root (path passed to load()).
    return pq.read_table("data/raw_orders.parquet")

The return value is registered into DuckDB (zero-copy for Arrow tables and DuckDB relations). Returning a str runs it as SQL; returning None means the function did its own work via context.connection.

Want to ingest from Postgres? Install psycopg yourself and let DuckDB pull it natively:

def model(context):
    context.connection.execute("INSTALL postgres; LOAD postgres;")
    context.connection.execute(
        "CREATE OR REPLACE TABLE raw_orders AS "
        "SELECT * FROM postgres_scan('host=… dbname=…', 'public', 'orders')"
    )

Loading SQL definitions from somewhere else

Models usually live in .sql files, but you can also feed them in directly from any source you can fetch yourself — a metadata table, a YAML registry, an HTTP endpoint. Unwind doesn't connect to anything; you bring the rows.

import unwind

rows = [
    {"name": "stg_users", "sql": "SELECT id, email FROM raw_users", "kind": "model"},
    {"name": "plus_one",  "sql": "{% macro plus_one(c) %}{{c}}+1{% endmacro %}", "kind": "macro"},
]
project = unwind.load_from_rows(rows, origin="catalog.sql_defs")

Reusing an existing DuckDB connection

Project.run() opens a fresh in-memory DuckDB by default, but you can pass your own connection — useful when you've already installed extensions, attached external databases, or configured secrets:

import duckdb, unwind

conn = duckdb.connect(":memory:")
conn.execute("INSTALL httpfs; LOAD httpfs;")
unwind.load("models/").run(connection=conn)
conn.execute("SELECT * FROM fct_warehouse_profitability").fetchall()  # still open

Test

uv run pytest
uv run ruff check
uv run ty check