DAGWorks Inc. provides an open-source platform designed to enable the development of reliable AI agents and applications. It offers an integrated development and observability experience tailored for AI projects, aiming to establish the foundations for Composable AI Systems. The platform emphasizes the importance of observability and introspection for creating dependable AI systems.
DAGWorks Inc. standardizes Python code for expressing data, machine learning, LLM, and agent workflows, pipelines, and applications through lightweight frameworks. This standardization facilitates collaboration and integration of observability and datastore needs with minimal effort, thereby accelerating production timelines and reducing total cost of ownership. The approach ensures maintainable code and manageable data processes, enhancing business performance by delivering reliable AI solutions.
DAGWorks Inc. offers two open-source projects:
- Burr, focused on AI applications (https://github.com/dagworks-inc/burr).
- Hamilton, focused on AI pipelines and workflows (https://github.com/dagworks-inc/hamilton), with additional information available at https://www.tryhamilton.dev.
Both Hamilton and Burr include self-hostable UIs, along with enterprise and SaaS options. Hamilton provides versioning, lineage/tracing, cataloging, and observability with a simple code modification. Burr offers tracing, observability, and persistence with a single line of code.