Dagger — Trivia & Interesting Facts¶
Surprising, historical, and little-known facts about Dagger.
Dagger was created by Solomon Hykes, the creator of Docker¶
Solomon Hykes, who created Docker in 2013 and transformed the software industry, left Docker (the company) in 2018 and co-founded Dagger in 2022. His thesis: the same containerization revolution that transformed how software is deployed should transform how CI/CD pipelines are built and run.
Dagger pipelines run identically on laptops and in CI¶
Dagger's core promise is that if a pipeline works on your laptop, it works in CI — and vice versa. This is achieved by running every pipeline step in a container using BuildKit (the same engine that builds Docker images). This eliminates the "works on my machine but fails in CI" problem for pipelines themselves.
Dagger switched from CUE to SDKs after user feedback¶
Dagger's initial version (v0.1-v0.2) used CUE, a configuration language created by Marcel van Lohuizen (formerly of Google). After user feedback revealed that learning a new language was a major adoption barrier, Dagger pivoted to language-native SDKs (Go, Python, TypeScript) that let developers write pipelines in languages they already knew.
The Dagger Engine is built on top of Docker's BuildKit¶
Dagger's execution engine uses BuildKit, the same low-level build system that powers docker build. This was a pragmatic choice: BuildKit already solved content-addressable caching, parallel execution, and cross-platform builds. Dagger adds a programmable API layer on top.
Dagger modules create a marketplace of reusable pipeline components¶
Dagger's module system lets developers share reusable pipeline components across languages. A Go module can call a Python module can call a TypeScript module, all within the same pipeline. This cross-language composability was inspired by the shortcomings of GitHub Actions marketplace entries, which are often poorly maintained and hard to compose.
The caching model gives Dagger dramatic speed advantages¶
Dagger caches every operation by content hash, meaning if inputs haven't changed, the operation is skipped entirely. In benchmarks, this has shown 2-10x speedups over traditional CI systems that re-run entire pipelines. The caching works locally too, so iterating on a pipeline on your laptop is nearly instant after the first run.
Solomon Hykes called CI/CD "the last frontier of containerization"¶
In his 2022 announcement of Dagger, Hykes wrote: "CI/CD is the last major component of the software stack that hasn't been containerized. Applications run in containers. Infrastructure is defined in code. But pipelines? They're still bash scripts glued together with YAML." This framing positioned Dagger as completing the container revolution.
Dagger raised $30 million in Series A funding¶
Dagger raised a $30 million Series A in 2023, demonstrating significant investor confidence in the idea that CI/CD pipelines need the same containerization treatment that applications received a decade earlier. The round was led by Redpoint Ventures, with participation from existing investors.
The GraphQL API was a controversial architectural choice¶
Dagger uses a GraphQL API as its core interface, with language SDKs generating clients against this API. This was controversial in the DevOps community where GraphQL is less common. The team argued that GraphQL's typed schema and introspection capabilities made it ideal for a cross-language tool API.
Dagger can replace Makefiles, shell scripts, AND CI configuration¶
Many teams use Dagger not just as a CI engine but as a replacement for Makefiles and shell scripts used in local development. A single Dagger pipeline handles make build, make test, make deploy, and the CI/CD configuration — eliminating the common pattern of maintaining the same logic in three different places.