Modern data lakehouse platform. Plugin-driven. Storage-agnostic.

Features

• Template-first projects — phlo init creates focused starters for CSV, REST APIs, dbt medallion projects, Sling replication, and observability demos
• Decorator-driven ingestion — phlo-dlt and phlo-sling register assets without hand-written Dagster boilerplate
• Type-safe quality contracts — phlo-pandera schemas validate data before it lands in managed tables
• Capability plugins — packages contribute services, CLI commands, assets, resources, catalogs, hooks, and Observatory surfaces through Python entry points
• Storage-agnostic data plane — Iceberg, Delta, ClickHouse, Trino, MinIO, RustFS, Nessie, and PostgreSQL can be composed as needed
• Operator surfaces — phlo-api, Observatory, MCP, PostgREST, Hasura, Superset, pgweb, and observability packages expose runtime state and actions
• Local-first operations — phlo services generates and runs the project stack through Docker or Podman

What It Looks Like

from pathlib import Path

import pandas as pd
import pandera.pandas as pa
from pandera.typing import Series

import phlo


class EventsSchema(pa.DataFrameModel):
    id: Series[int] = pa.Field(ge=1)
    name: Series[str]
    value: Series[int] = pa.Field(ge=0)

    class Config:
        strict = True
        coerce = True


@phlo.ingest.dlt(
    table_name="events",
    unique_key="id",
    validation_schema=EventsSchema,
    group="demo",
    freshness_hours=(1, 24),
)
def events(partition_date: str):
    return pd.read_csv(Path("data/events.csv"))

phlo.ingest is the provider-neutral ingestion namespace. Use phlo.ingest.dlt(...) for Python, REST, and DataFrame sources; phlo.ingest.sling(...) for replication; or phlo.ingest.provider("name") for third-party ingestion providers. Existing @phlo.ingestion(...) workflows remain supported as a DLT compatibility alias.

Prerequisites

• uv — Python package manager
• Docker — Container runtime

Quick Start

# Install with default plugins
uv pip install phlo[defaults]

# Initialize a runnable local starter
phlo init my-project --template csv-batch
cd my-project

# Generate service configuration, start services, and materialize
phlo services init
phlo services start
phlo materialize dlt_events --partition 2026-05-04

Documentation

Full documentation source lives under docs/index.md. The published site is generated with pymdx, matching the GitHub Pages workflow:

pymdx generate src/phlo --docs docs --output docs-site
pymdx build docs-site

Primary entry points:

• Installation Guide
• Quickstart Guide
• Core Concepts
• Developer Guide
• Plugin Development
• Workflow Development
• CLI Reference
• Configuration Reference
• Operations Guide

Development

uv pip install -e .    # Install Phlo in dev mode
make check             # Lint, format, typecheck, and test (parallel)

# Services
phlo services start    # Start infrastructure
phlo services stop     # Stop services
phlo services logs -f  # View logs

# Individual gates
uv run ruff check .    # Lint
uv run ruff format .   # Format
uv run ty check        # Typecheck
uv run pytest          # Test

Architecture

Phlo is a monorepo of composable packages — install only what you need:

Layer	Packages
Orchestration	phlo-dagster
Ingestion	phlo-dlt, phlo-sling
Quality	phlo-pandera
Transforms	phlo-dbt
Table formats	phlo-iceberg, phlo-delta, phlo-clickhouse
Infrastructure	phlo-traefik, phlo-postgres, phlo-oauth2-proxy
Storage	phlo-minio, phlo-rustfs
Catalog	phlo-nessie, phlo-openmetadata
Query	phlo-trino
Observability	phlo-otel, phlo-clickstack, phlo-grafana, phlo-prometheus, phlo-loki, phlo-alloy, phlo-alerting
UI	phlo-observatory, phlo-pgweb, phlo-superset
API	phlo-api, phlo-mcp, phlo-hasura, phlo-postgrest
Dev/Test	phlo-testing