Choosing a sandbox runtime for long-lived coding agents
How we evaluated sandbox providers for coding agents, what we chose for now, and the trade-offs behind that decision.
Start locally without complicated setup hassle
- ZenML is available as a simple pip package that lets you run and track pipelines locally.
- ZenML integrates with your orchestration layer of choice, avoiding having to learn different paradigms for dev, staging, and prod.
- ZenML integrates with your orchestration layer of choice or can be extended with your own orchestration service.
Abstract away infrastructure complexity
- Most orchestrators assume some form of infrastructure knowledge to use them maximally — ZenML abstracts that complexity away.
- ZenML separates infrastructure setup like Docker building from the application logic, and automates the tedious parts.
- ZenML focuses on the handovers between MLOps Engineers, ML Engineers, and Data Scientists.
Switch between orchestrators depending on your context
- You can switch between different orchestration services with a single click — from dev to staging to production.
- The more engineering-minded in the team still retain control over their productionalization because the framework is extensible.
- ZenML handles the pain of packaging your code into Docker to be deployed to your orchestration service of choice.
Explore in Detail What Makes ZenML Unique
| Feature | ZenML | Prefect |
|---|---|---|
| ML-Centric Design | Yes Purpose-built for machine learning workflows | Not supported General-purpose workflow orchestration |
| Experiment Tracking | Yes Built-in experiment tracking and comparison | Not supported Relies on external tools for experiment tracking |
| Model Registry | Yes Integrated model registry for versioning and deployment | Not supported No built-in model registry |
| Hyperparameter Tuning | Yes Native support for hyperparameter tuning | Not supported Requires custom implementation |
| ML Framework Integration | Yes Deep integration with popular ML frameworks (scikit-learn, TensorFlow, PyTorch) | Yes Supports integration with various ML frameworks |
| Pipeline Definition | Yes Clean and intuitive pipeline definition using Python decorators | Yes Flexible pipeline definition using Python |
| Task Parallelism | Yes Supports parallel execution of pipeline steps | Yes Enables parallel execution of tasks |
| Data Versioning | Yes Integrates with data versioning tools for reproducibility | Not supported No built-in data versioning support |
| Cloud Integration | Yes Built-in support for deploying pipelines to cloud platforms | Yes Supports deployment to various cloud platforms |
| Workflow Scheduling | Yes Schedules and triggers ML pipeline runs | Yes Robust scheduling and orchestration of workflows |
Code comparison
ZenML and Prefect side by side
from zenml import pipeline, step
from zenml.integrations import mlflow
@step
def preprocess_data(data):
# Preprocess the data
preprocessed_data = ...
return preprocessed_data
@step
def train_model(preprocessed_data):
# Train the model
model = ...
return model
@step
def evaluate_model(model, test_data):
# Evaluate the model
metrics = ...
mlflow.log_metrics(metrics)
return metrics
@pipeline
def ml_pipeline(data, test_data):
preprocessed_data = preprocess_data(data)
model = train_model(preprocessed_data)
metrics = evaluate_model(model, test_data)
# Run the pipeline
ml_pipeline(data, test_data)from prefect import task, Flow
from prefect.tasks.ml import mlflow
@task
def preprocess_data(data):
# Preprocess the data
preprocessed_data = ...
return preprocessed_data
@task
def train_model(preprocessed_data):
# Train the model
model = ...
return model
@task
def evaluate_model(model, test_data):
# Evaluate the model
metrics = ...
mlflow.log_metrics(metrics)
return metrics
with Flow("ml_pipeline") as flow:
data = ...
test_data = ...
preprocessed_data = preprocess_data(data)
model = train_model(preprocessed_data)
metrics = evaluate_model(model, test_data)
# Run the pipeline
flow.run()01.
ML-Centric Design and Features
ZenML is purpose-built for machine learning workflows, offering a range of ML-specific features such as experiment tracking, model registry, and hyperparameter tuning, which are not natively available in Prefect.
02.
Intuitive Pipeline Definition and Reusability
ZenML provides a clean and intuitive way to define ML pipelines using Python decorators, making it easy to encapsulate and reuse pipeline components across different projects and experiments.
03.
Deep Integration with ML Frameworks and Tools
ZenML offers deep integration with popular ML frameworks like scikit-learn, TensorFlow, and PyTorch, as well as seamless integration with other MLOps tools, enabling end-to-end orchestration of ML workflows.
04.
Reproducibility and Governance
With built-in support for data versioning and lineage tracking, ZenML ensures reproducibility, traceability, and governance of your ML pipelines, facilitating compliance and collaboration.
05.
Extensible Architecture and Customization
ZenML's extensible architecture allows you to customize and extend its functionality to fit your specific ML workflow requirements, providing flexibility and adaptability as your needs evolve.
Outperform Orchestrators: Book Your Free ZenML Strategy Talk
Orchestrator Showdown
Explore the Advantages of ZenML Over Other Orchestrator Tools
ZenML vs Apache Airflow
ML-optimized workflow management. Enhance scalability and usability with comprehensive features designed for ML pipeline orchestration.
ZenML vs Argo Workflows
ZenML is an open-source alternative to Argo Workflows for ML pipelines with built-in metadata, lineage, and reproducibility
ZenML vs Dagster
Dagster alternative: Streamline ML ops with intuitive pipelines. Seamless integrations and experiment tracking for efficient ML workflow management.
ZenML vs Databricks
Databricks alternative: Flexible ML orchestration without vendor lock-in. Accelerate ML with lightweight, adaptable workflows across multiple clouds.
ZenML vs Flyte
Flyte alternative: Agile ML pipeline orchestration. Accelerate workflows with intuitive tools, seamless MLOps integration, and rapid iteration.
ZenML vs Kedro
Kedro alternative: Scalable, user-friendly ML framework. Streamline operations with robust features for efficient project management and deployment.
ZenML vs Kubeflow
Kubeflow alternative: Lightweight ML pipeline management. Simplify MLOps with flexible, user-friendly workflows across various environments.
Broaden Your AI Orchestration Understanding with ZenML
Introducing the new Kitaru: from production traces to repeatable evals
The new Kitaru turns production traces you already collect into replayable test scenarios, expert-reviewed cohorts, and evaluators you can run on every change.
Your GPUs Are Everywhere. Your Robot-Learning Loop Shouldn't Be.
Robotics compute is spreading across clouds and clusters. Learn how one portable pipeline layer can keep the robot-learning loop reproducible.
Unlock the Full Potential of
Your ML Workflows with ZenML
Experience the power of ZenML's ML-centric design and features tailored for machine learning workflows Define and orchestrate your ML pipelines with ease using ZenML's intuitive and reusable components Seamlessly integrate ZenML with your favorite ML frameworks and MLOps tools for end-to-end workflow management Ensure reproducibility, traceability, and governance of your ML pipelines with ZenML's built-in lineage tracking and versioning capabilities