•Roadmap
Arbiter is a new project and we are just warming up. We have a lot we want to achieve and this page will walk you through what we are planning to build in the coming year.
This is not an exhaustive list nor is it a promise that these features will be built, but it gives a glimpse into the vision we have for this project.
•Embracing Large Language Models(LLMs)
At Arbiter, we are passionate about the potential of large language models. These models have already shown their incredible capabilities in a wide range of applications, including natural language processing, code generation etc. However, we are particularly excited about their potential application in low-code platforms.
Low-code platforms are designed to make it easier for users with varying levels of technical expertise to create applications without writing extensive lines of code. With the help of large language models, these platforms can become even more powerful and intuitive, allowing users to create complex applications with ease.
We are dedicated to exploring the potential of large language models in low-code platforms, and we believe that these models can help to revolutionize the way that users create applications. By providing more powerful and intuitive tools, we can help to make software development more accessible and efficient, enabling users to create innovative solutions. We have begun developing a Text-To-Integration feature and an AI function builder to streamline the processes of creating HTTP connectors and function processors. This is just the start of our endeavor to fully utilize LLMs and deliver an exceptional user experience.
•Human tasks processor
We plan to build a mechanism by which human input will be added to the workflow execution. Execution will wait until the user fills in the required input, and after completing the form will continue to the next node,
•SQL connector
We plan to create a SQL connector and support the most popular SQL databases - Postgres, MySQL, MSSQL, etc.
•AI/ML processors
In the future, we want to introduce an AI/ML processor. Arbiter will be able to work with machine learning and deep learning models which will be uploaded by the user. We plan to support Tensorflow, R, NumPy, Pandas, Keras, PyTorch, Matplotlib, et al.
•Error handling
In the future, users will be able to handle errors with Rule node. We'll introduce a meta field error which will be available in every node and will represent errors from the previous node.
•Binary data support
Currently, Arbiter does not support raw binary data.
•User-defined tests
Arbiter will be able to execute user-defined tests before workflow deployment.
•Give Feedback
If you have an unaddressed need in this roadmap then let us know! If you'd like, you can provide feedback.