
OpenAI
by Appian Corporation
Description
Overview The OpenAI Connected System allows users to send prompts and receive AI generated responses, whether that be images (DALLE-3), audio translation/transcription (Whisper), or text completions (GPT-4). These responses can be edited, updated, or expanded upon as desired. Users can also fine-tune (custom train) a model based on Record data to fit their workflow, summarize Record content, check for content-policy violations, edit images, create vector embeddings, along with much more.
Features
Key Features: ChatGPT Completion – Augment your workflow with ChatGPT's enhanced text generation capabilities. ChatGPT Function Calling — Allow GPT to select a function to call and provide arguments to call that function to back your user's query with live data Audio Transcription/Translation – Transcribes/translates audio into the input language Create DALLE Image – Given a prompt and/or an input image, the model will generate a new image Edit/Extend DALLE Image – Creates an edited or extended image given an original image and a prompt Create Variation of Image – Creates a variation of a given image Create JSON Lines File – Creates a JSON Lines File from Appian data Create Fine-Tuning Job – Creates a job that fine-tunes a specified model from a given dataset
Release notes
Version 2.0.1:
- Applied bug fixes
Version history (12)
2.0.1
Version 2.0.1:
- Applied bug fixes
2.0.0
Version 2.0.0:
- Security vulnerability fixes
1.2.6
Version 1.2.6:
- Security vulnerability fixes
1.2.5
- Search library update
1.2.4
- CVE patches
1.2.3
- CVE patches
- Updated to OpenAI API to version 2.0.0
1.2.2
Now compatible with OpenAI's latest API improvements, including new models (gpt-4-0613 & gpt-3.5-turbo-16k) and Function Calling. Learn more about these enhancements here: https://openai.com/blog/function-calling-and-other-api-updates.
1.2.1
Patches language bug in the OpenAI ChatGPT component.
1.2.0
OpenAI ChatGPT Component Plugin Support: Bring the unmatched power of NLP and conversational AI to Appian workflows with the OpenAI ChatGPT component plugin. This connected system securely stores OpenAI authentication credentials and is required to use the ChatGPT component plugin.
1.0.0
ChatGPT is now available through the /chat/completions endpoint. Build ChatGPT into your Appian Application or use it's enhanced text generation capabilities inside your workflow. Audio to text transcription and translation now available through the OpenAI Whisper API. Use this functionality with the new /audio/transcriptions and /audio/translations endpoints UI update. Use ”Specify values for each input” to update the parameter values individually and use “Define all values with a single expression" to view all parameters and their descriptions at once. "Stream" parameter bug fix. Streaming content is not available and this parameter has been removed.
1.1.0
ChatGPT: ChatGPT is now available through the /chat/completions endpoint. Augment your workflow with ChatGPT's enhanced text generation capabilities. Audio Transcription/Translation: Audio to text transcription and translation now available through the OpenAI Whisper API. Use this functionality with the new /audio/transcriptions and /audio/translations endpoints. UI update: Use ”Specify values for each input” to update the parameter values individually and use “Define all values with a single expression" to view all parameters and their descriptions at once. Bug Fix: "Stream" parameter bug fix. Streaming content is not available and this parameter has been removed.
1.0.0
Sample app hotfix.
Resources
- Download
OpenAI Documentation.pdf
Documentation