Projeyle

Projeyle is a modular software platform designed to orchestrate multi-step computational workflows across heterogeneous computing environments. It provides a declarative workflow language, a plugin-based architecture for data sources, processing steps, and execution backends, and built-in support for provenance, versioning, and reproducibility. Workflows in Projeyle are described as directed acyclic graphs of tasks, with explicit dependencies and retry policies.

The architecture centers on the Projeyle Engine, which schedules and executes tasks, and the Core library, which implements common workflow primitives. The system uses adapters to connect to local workstations, high-performance clusters, and cloud resources, and relies on containerization to isolate environments. Pipelines can be defined in YAML or JSON, and can be executed incrementally with caching to avoid reprocessing unchanged steps. Projeyle emphasizes portability of pipelines across platforms and provides mechanisms for recording provenance data, including inputs, software versions, and parameter values.

Projeyle originated from the Projeyle Consortium in the early 2010s as an attempt to unify disparate workflow tools and improve reproducibility in computational science. The project is community-governed, with a reference implementation released as open source. Since its introduction, several ecosystem plugins have been released for data formats, databases, and visualization.

Usage and reception: Projeyle is used in bioinformatics, environmental modeling, and data science research. Advocates point to its portability and reproducibility; critics note the learning curve and the complexity of configuring environments and plugins. The platform competes with other workflow systems but remains relevant in settings that require end-to-end provenance and scalable execution.