A: This is something you can instruct the Assistant to avoid. By adjusting the Assistant’s prompt, you can have full control over how the Assistant behaves and responds. If there is a concern about how the Assistant will respond in a particular situation, simply include instructions on how you’d like it to proceed.

A: Conversations are private to the user that sends them and are not shared. The user has the ability to rename and export any previous conversations if they would like to share them with others. It is on the roadmap to create an analytics dashboard to share insights from usage, but individual messages will not be disclosed.

A: Yes, Nectir AI supports over 50 languages, including English, Spanish, French, Portuguese, Russian, Japanese, Chinese, and Arabic.

A: Nectir AI offers one-to-one conversations between the user and the Assistant. Any user with access to a workspace can utilize their assistants at any time, as often as they would like.

A: Speech to text functionality is on the roadmap, and something we want to build soon.

A: Currently, Nectir AI can only be trained with documents or images. To train your assistant on a video, we recommend you upload a transcript of the video for similar results.

A: Yes! The AI Assistant can access the information from these documents; however, the user will not be able to see where the information is being pulled from. Non-student-facing materials such as instructional guides, student feedback, and assessment blueprints can help train the AI to provide more informed and contextually relevant support to both educators and students.

A: The Assistant cannot access the internet directly, so it does not have real-time access to information from the web. However, the AI model behind the Assistant has been trained on a diverse corpus of information, including books, articles, websites, and other written material, and optimized for accuracy across a wide range of topics.

Learn more about setting your Assistant’s knowledge scope here to understand how you can limit the information your AI Assistant has access to from the LLM.

A: Both! It is up to you how you would like to set up a Nectir AI Assistant for your class. Assistants can be designed to address a variety of problems inside and outside the classroom.

For example, a Course Assistant designed to support students in their coursework can help answer student questions 24/7, enabling them to continue following along during a lesson.

A: User data is securely stored and never shared with third parties. While personally identifiable information may be stored for moderation and safety purposes, only anonymized data is used for product improvements.

A: We leverage Microsoft Azure’s sandboxed endpoint of GPT-4 to power Nectir AI Assistants. That way, no data is shared with third parties or used to train any models. This makes our platform more secure than going directly to ChatGPT or Claude.

A: Since FERPA compliance is up to the interpretation of each individual school, no tool can claim to be fully "FERPA-compliant." With that said, Nectir AI has been able to comply with all major data safety and privacy requirements across dozens of campuses related to FERPA.

A: Only resources that are publicly available, open-source (such as OER), or have the consent of the owner can be used to train the Assistant for now. We do not have any partnerships with publishers yet, but this is something we're working on.

Since only instructors have the ability to upload training documents, it is their responsibility to not share copyrighted material with the Assistant or violate any publishing licenses.

Note: Your uploaded training materials are securely stored and never shared or used to train AI models. Materials will only be used within the scope of your AI assistant.

A: Nectir securely stores all uploaded materials and never shares information with third parties. We also host our own language model endpoints so that none of your information is shared with third-party companies like OpenAI.

A: We use OpenAI’s GPT-4o (Omni) model.