Featured Article: Using Dato’s GraphLab Create to Perform Record Linkage Efficiently by Nick McClure (Guest Blog)

Ultra-Fast Data Analytics

Wrangle TBs of data at interactive speed

Use SFrames for fast data preparation and data engineering for machine learning. Connect with:

  • SQL databases
  • Hadoop HDFS
  • Spark RDD
  • Apache AVRO
  • AWS and S3
  • Python Pandas.DataFrame
  • CSV files
  • JSON files

Slice and dice all your data, engineer features, explore and visualize its statistical properties, fast.


We optimized SFrames and open sourced it for the data science community

SFrame is an out-of-core, columnar data structure. So, it’s not limited by main memory, letting you handle all your data at once, even on your laptop.



Download GraphLab Create™


Best-in-Class Predictive Modeling

GraphLab Create™ offers a deep library of Python machine learning APIs and toolkits

We provide high performance algorithms for:

  • Recommenders
  • Data matching
  • Deep learning
  • Sentiment analysis
  • Churn prediction
  • Personalization
  • Object recognition
  • Topic modeling
  • Classification
  • Clustering
  • Regression
  • Graph analytics
  • Neural networks
  • Matrix factorization
  • Image processing
  • Text analytics


Leverage the most effective machine learning algorithms for your task. Optimized and tuned.

Tackle data mining, big data analytics and predictive modeling tasks with the most advanced machine learning out there. Combine our APIs and toolkits with scikit-learn and all your existing Python code, easily integrating with your existing stack. And, our distributed auto-tuning makes it all simpler.

import graphlab as gl
data = gl.SFrame.read_csv('my_data.csv')
model = gl.recommender.create(data,
recommendations = model.recommend(k=5)




Download GraphLab Create™


Production-Ready Data Science

RESTful deployment is easy, fast, and robust

One line of code deploys all your Python machine learning models as REST APIs, including GraphLab and scikit-learn code. Building intelligent apps and integrating with other dev/ops teams is fast and easy.

Dato Predictive Servers take your models to production:

  • Millisecond response time
  • Elastic, robust, cached, load-balanced web service, on AWS or on-premises
  • Incorporate business logic and arbitrary Python code
  • Monitor, manage, and update deployed models with no downtime


Scale with distributed machine learning

Dato Distributed Servers distribute all your tasks:

  • Distributed machine learning
  • Any Python code, including scikit-learn
  • On-premises or on AWS
  • Deeply integrate with Hadoop, Yarn and Spark
import graphlab as gl
env = gl.deploy.environment.EC2(
ps = gl.deploy.predictive_service.create(
ps.add('recommender', model)




Download GraphLab Create™


Our Customers



At Pandora, we use and investigate a lot of new machine learning and big data tools, GraphLab Create™ helps us iterate on ideas for new product features faster and at large scale.Oscar Celma, Director of Research, Pandora


I've found that Dato deeply understands the needs of the data scientist. The ease of use and scalable performance, which is not limited by the memory of the machine, are allowing us to innovate and advance at an astonishing pace. Andrew Bruce, Senior Director of Data Science, Zillow


GraphLab Create™ meets our goals for speed and scale. Couple the power of the GraphLab Create™ platform with an extremely responsive support team and we can deliver personalized data insights to our customers.Jonathan Seidner, CTO, Crosswise
Hotel Tonight


Within 30-minutes of installing Graphlab Create, I had built a text classifier with 90% accuracy analyzing hotel reviews from our customers. Chas Lemley, Software Engineer, Hotel Tonight