Over the last weeks we’ve spent a number of blog posts explaining the ways in which AgilePoint will support public models like Gemini, OpenAI, Claude and others. In this installment, we’re happy to announce support for other models that leverage the OpenAI Compatible APIs and spec.
Why Is This Important?
For a couple of reasons mainly. For one, having this generic support for models that support OpenAI spec allows AgilePoint a wide range of models that can easily be integrated into your next application. Think of this as the LLM equivalent of the REST protocol connector, with the added benefit of abstraction and avoiding the technical details.
Secondly, this connector allows AgilePoint to meet the growing demand for locally-deployed LLMs that are trained not on the general knowledge of the internet, but your organization’s specific context!
As with any AI platform, our support includes both process activities and AI Control Towers.
Process Activities:
- Analyze Image
- Chat Completion
Control Tower Activities:
- Analyze Image
- Analyze Image – Adapt Flow
- Analyze Image – Ad Hoc Task
- Analyze Image – Initiate Sub process
- Analyze Image – Send Notification
- Chat Completion
- Chat Completion – Adapt Flow
- Chat Completion – Ad Hoc Task
- Chat Completion – Initiate Sub process
- Chat Completion – Send Notification
Note: Please note that this video was recorded while this feature was in beta version. Some of the labels and icons might change by the time it goes GA but the general concept of configuring these activities would remain the same.