DEVOPSdigest

RapDev released Digi, an open-source Coding Assistant for ServiceNow, with free access for the ServiceNow community.

Digi grants ServiceNow developers a GitHub co-pilot experience on the platform, making code generation and debugging effortless, while continuing to drive innovation within the ServiceNow ecosystem.

"LLMs have been in the wild. The tricky part is enabling their output to create meaningful change in a technical environment, in a controlled and configurable way. I have never been prouder of a team of engineers. They delivered the foundation of an open-source AI future for the entire ServiceNow ecosystem. This is just the beginning!", said Kyle Brueckner, Engineering Manager at RapDev.

By offering a flexible and configurable framework to leverage AI in modern operations, developers can benefit from spending less time writing code, and more time automating testing and debugging to innovate faster with ServiceNow. Live on GitHub, ServiceNow developers can now access key features:

- Connect to popular models or AI engines like Codex, Bard, or private Starcoder stacks.

- Customize prompts, prefixes, and postfixes for messages sent to the AI engine

- Complete traceability for every AI action performed on the platform.

- Validate changes to ensure modifications to your environment are accurately assessed before implementation.

A simple user-interface allows a ServiceNow developer to submit a statement, which in turn returns an AI-generated block of code that can be easily inserted into any ServiceNow development scripts. A developer can review, insert, test, and accept the block of code within seconds, saving them minutes or even hours of manual development.

The road ahead for Digi

- Pre-trained data collation: Streamline the preparation of structured pre-training data by leveraging relevant and publicly available information from your ServiceNow instance.

- Starcoder in a box: Offer Infrastructure-as-Code (IaC) for a plug-and-play Starcoder stack, enabling you to securely train sensitive data.

- Test-Driven Development: Enable Digi to build solutions on the platform based on ATF (Automated Test Framework) tests until the tests pass.

- AI Actions: Extend Digi's use cases beyond code generation and harness its vast data resources to provide contextual insights for workflow creation, ticket handling, and real-time incident analysis.