Daft is a Dataframe library optimized for quick experimentation and iteration on Notebooks.
Daft runs anywhere - on your local machine or on a Ray cluster when you need more horsepower to work on all that data.
Daft is free and open-sourced! You can use it alongside other open-sourced Python tools such as Numpy, Pillow and more for your data processing needs.
Daft Dataframes are a simple, elegant abstraction on top of your data.
You have both the expressiveness of SQL, but also access to native Python transformations on columns that contain images, vector arrays and more.
All of this seamlessly runs on the Eventual Cloud Platform without any infrastructure management on your part.
Eventual uses open data formats such as Apache Parquet and Apache Iceberg to store data in your cloud accounts. We provide integrations with common cloud data storage platforms such as AWS S3, PostgreSQL and Snowflake so that you never have to worry about writing efficient Data I/O or serialization code.
Daft is built for performance and scalability.
When running experimentation and data exploration on a single machine, it outperforms frameworks such as Pandas. When run in the cloud, Daft's intelligent scheduler and optimizers can scale workloads to Petabytes of data for distributed computations.
Daft is built for high-performance computing and for first-class GPU support to accelerate data processing for many computer vision/machine learning/deep learning workloads.
Jay graduated from Cornell University where he did research in Machine Learning and Computational Biology at the Yu and Danko labs. He is from Singapore, and was a tank platoon commander in the Singapore army as well as the Head of Talent Acquisition at Shopback. Jay was the founding engineer of the Machine Learning (ML) platform team at Freenome, building out a platform to detect colorectal cancer from genomic data, and later joined Lyft Level 5 as a senior engineer focusing on ML infrastructure for distributed deep learning.
Eventual provides managed infrastructure for all your compute, storage and data science environment needs, including managed DaFt deployments that can scale to any size your organization requires.
Get in touch today to chat with us about your data needs/requirements!