Scale-Out Delivery Platform→
Complexity is inevitable. Tame it and gain your competitive advantage.
Capabilities
Pipelines→
Build the CI/CD workflows you need.
Test Engine→
Remove flaky tests, split tests, optimize performance.
Package Registries→
Speed up builds, lock down security.
Mobile Delivery Cloud→
Supercharge your mobile app delivery.
Bring your own compute
Self-hosted agents Self-hosted Kubernetes Self-hosted AWS Self-hosted Google Cloud
Hosted compute
Mac hosted agents Linux hosted agents
Replace Jenkins
Shift your software delivery into high gear.
Workflows for AI/ML
Tame the chaos of model delivery today to prepare for 10× tomorrow.
Testing at scale
Deliver quickly and confidently, at any scale.
Monorepo mojo
Scale your monorepo without scaling your problems.
Webinars
Register for upcoming sessions or watch on-demand recordings.
Blog
Insights from the team building Buildkite.
Case studies
Real-world customer success stories.
Events
Join us at industry conferences and community meetups.
Changelog Plugins Releases

About

The company behind Scale-Out Software Delivery.

Careers

Join the team!

pipeline.yaml

name: "Join Buildkite"
description: "Fetch all jobs, filter by relevance, and apply"
steps:
  - command: "curl -X GET 'https://boards-api.greenhouse.io/v1/boards/buildkite/jobs' \ -H 'Content-Type: application/json' \ -v"
    label: ":terminal: Fetch latest jobs"

Press and news Brand assets

Follow Buildkite

Resources
/
Webinars
/
How Culture Amp uses DevOps techniques for reproducible machine learning

How Culture Amp uses DevOps techniques for reproducible machine learning

Learn how Culture Amp adapts conventional CI/CD to the complexities of machine learning.

Watch the webinar

Recorded on: February 1, 2024

Length: 29 minutes

In this webinar (originally presented at Buildkite's Unblock conference), Andy covers what machine learning is and the complexities that require special treatment in CI/CD pipelines. In particular, Andy talks about:

Using natural language processing (NLP) models to analyze employee survey responses.
Differences to conventional pipelines due to the scale of data and number of parameters.
Tracking the code version and parameters that produce a model.
Integrating with third-party tools like AWS Batch, Athena, and SageMaker with Buildkite's API.
Including human inspection before making deployment decisions.

Related webinars

Buildkite presenter Mike Morgan with Rippling team members discussing reducing infrastructure cost through moving CI to AWS spot instances

March 27, 2024

40 minutes

How Rippling reduced cost and improved developer experience by moving CI to spot instances

Learn how Rippling optimizes CI/CD with Buildkite and AWS spot instances, including cost evaluation, managing spot challenges, ensuring fast builds, and maintaining developer experience.