Verawood delivery for AI Extensions project - documentation

[sarina]

@felipemontoya @e0d - document what our deliverables for Verawood are so we can make product release notes & product marketing

[felipemontoya]

I was contacted by @crathbun428 about this. I will copy it here to track it.

1. User Test Cases for Community Release Testing (due to me by 4/20) to set the community up for testing your feature as part of the Community Release Testing process. It is up to you and the delivery team what components of your feature are important for the community to stress test if at all. I've created a Release Testing Sheet Template linked here for you to fill in to make it easy for us to compile all the new tests in the formal testing sheet. I left a few filled in rows at the top just to give you a sense of what belongs in each column, and I've hidden a few columns you don't need to worry about. I'm also linking a resource here about writing a good test case in case it is at all helpful.

2. Product Release Notes (due by May 7) to inform and get non-technical users excited about the feature's inclusion in the upcoming release. These don't need to be thorough how-to's, more a quick overview of what the feature is, where it is, and what it's value is to the user/how it can be used. These can link out to the feature's documentation for those who want to learn more detailed information about how the feature works. Here is a link to some recent product release notes in case you find reviewing these helpful. Please add your Product Release Notes for your feature(s) to the appropriate page linked here: https://openedx.atlassian.net/wiki/spaces/OEPM/pages/6221004801/VERAWOOD+-+DRAFT+Product+Release+Notes

3. Operator Release Notes (due by May 7) to inform the operator who will be setting up and/or managing the instance of the feature's addition along with any critical they need to know about how to enable, configure, or disable the feature. Please add any information you need to add to the Operator Release Notes about your feature to this wiki page, here: https://openedx.atlassian.net/wiki/spaces/COMM/pages/5331222534/Verawood+-+Operator+Release+Notes . Please note: You'll see there's a space at the bottom of this page to add any instructions you would like the Release Manager in the Build-Test-Release working group to know so that they can make sure your feature is configured appropriately for the release and for the release testing sandbox.

I'll work on 2 and 3 while we solve the sandbox for testing with the BTR.

[sarina]

@felipemontoya I've been having fun writing documentation with Claude and could probably take a stab at this, but it'd be really helpful if you could share with me any spec documentation, PRs/ADRs that describe changes, etc that I could feed in (I have a pretty-well tailored memory on generating Open edX docs)

[felipemontoya]

Thats a very good use of claude.

Lets see, here are the main ideas I would like to say for operator notes:

Once you install it, the feature is enabled by default, but in order to use it, you need to manually activate for every course or activity using the /admin/ console.
Link: https://docs.openedx.org/projects/openedx-ai-extensions/en/latest/readme.html#id1

The basic requirement is that you set some keys in the config.yml if you are using tutor. Make it reach the settings if you dont use tutor.

Necessary:

    AI_EXTENSIONS:
      openai:
        API_KEY: "sk-proj-your-api-key"
        MODEL: "openai/gpt-5-mini"
      <provider>:
        MODEL: "ollama/yourmodelname"   
        OTHER_CONFIG_VALUES: "url of your model perhaps"

    PLUGINS:
      - openedx-ai-extensions

Optional:

    AI_EXTENSIONS_ENABLE_LLM_CACHE: true

General guide:
Usage guide: https://docs.openedx.org/projects/openedx-ai-extensions/en/latest/quickstarts/usage_guide.html

If you're using this in combination with aspects, you can run the aspect load assets pipeline and it will create dashboards for AI-extensions usage.

---

For product notes, I would like to say something along these lines:

https://docs.openedx.org/projects/openedx-ai-extensions/en/latest/quickstarts/usage_guide.html

Main features:
- Unit content summarization with AI: https://docs.openedx.org/projects/openedx-ai-extensions/en/latest/quickstarts/usage_guide.html#lms-example-content-summary
- Conversational chat with learners in the context of the course content with context awareness from the AI.
- AI Assistant for course authors to write questions based in the content into content libraries: https://docs.openedx.org/projects/openedx-ai-extensions/en/latest/quickstarts/usage_guide.html#studio-example-library-question-assistant
- Flashcards experience. Using AI to craft anki type cards in the learning experience.
- Badge generation using AI and the DCC project when installing a separate plugin: https://github.com/eduNEXT/openedx-ai-badges

For developers:

• the plugin has some experiences using AI, but it is highly customizable and extendable. It is designed so developers can take advantage of it and build many other experiencies. Case in point: https://github.com/eduNEXT/openedx-ai-badges
• link to docs for devs: https://docs.openedx.org/projects/openedx-ai-extensions/en/latest/quickstarts/extension_guide.html

---

Is this what you had in mind?
Also, I updated the usage docs and asked for your review here: #214