Source

• Text based features:
  • AI: ChatGPT text generation, generate text using GPT-3 of OpenAI.
    • The GPT-3 text generation delivers 1 result by default, however the advanced settings allow for as many results as you want and you can even choose the model of data to use for optimal results.
  • Translate text.
• Search content based features:
  • Help to share info (Location, Task, Table, GIF, and every sort of thing).
• Media features:
  • AI: Image generation, generate an image based on a description given by the user.
  • AI: Speech-to-text, audio transcription directly to the document.​​​​

Source

Confluence AI use:

Source

Microsoft Loop’s AI:

• Text based features:
  • Smart Summary: AI-powered summarization of documents or meetings into key insights and highlights.
  • Quickstarter: AI-powered templates to help you get started with common projects and tasks.
• Search content based features:
  • Automatic Suggestions: AI-powered suggestions for relevant content and actions based on your previous activity.
  • Content Classification: AI-powered classification of content into relevant categories for easier organization.
  • Intelligent Search: AI-powered search function that understands the context of your queries and provides relevant results.
  • Adaptive Cards: AI-powered card design that adapts to the user's preferences and environment.
• Media features:
  • Automatic Transcription: AI-powered transcription of audio and video recordings into text format.

Source

Most recurrent functionalities:

First approach:

Still following our categories defined at the beginning, we can establish these stats :

• 17 features are texts based features to help the users to go faster
• 13 features are about content research / content related proposition
• 3 features are about media transcription.

So we can see that the most common features based on these data are the one used to help the user in his tasks (writing / editing fast) and then the one to help him search content.

The media features are minor, but still have an interest : The NextCloud use of AI images generation is in fact a fast-pass for the user to get what they want for their content, like the text based features.

Conclusion:

The first thing we can say about which features is a priority, based only on the ones we found there, is that the priority is on making functionalities to help the user create content faster. The functionalities that aim to help on the search of content are important to consider too.

We have to consider users opinions on the question, to sort which features is the most demanded / useful.

Second approach – Analysis made with the help of ChatGPT 4:

Here we grade the features based on two criteria, importance of the feature and difficulty to implement it.

First, the grading of the most valuable integrations from each category based on their importance, with 10 being the most important:

1. Text-based features:

   • Language translation: 9
   • Grammar and orthography correction: 8
   • Automatic article writing from a title: 7

2. Content search and organization features:

   • AI-powered search with natural language processing: 9
   • Personalized content recommendations: 7
   • Automatic content tagging for organization: 8

3. Media transcription features:

   • Speech-to-text (audio transcription): 8

In this grading, language translation and AI-powered search with natural language processing have the highest scores, as they have broad applicability and can significantly improve the efficiency and effectiveness of user interactions with the software. Grammar and orthography correction, automatic content tagging, and speech-to-text also score highly due to their widespread usefulness and potential to enhance user experience. Automatic article writing and personalized content recommendations, while valuable, have slightly lower scores as their relevance may be more context-specific or depend on user preferences.

Finally, the grading of the integration difficulty of the most valuable features, with 10 being the easiest to implement and 1 being the most difficult:

1. Text-based features:

   • Language translation: 7
   • Grammar and orthography correction: 8
   • Automatic article writing from a title: 6

2. Content search and organization features:

   • AI-powered search with natural language processing: 5
   • Personalized content recommendations: 4
   • Automatic content tagging for organization: 7

3. Media transcription features:

   • Speech-to-text (audio transcription): 6

In this grading, grammar and orthography correction and automatic content tagging are relatively easier to implement, as they rely on well-established technologies and can be integrated with existing tools and services. Language translation and speech-to-text are moderately difficult to implement due to the complexity of human language and potential challenges with accuracy and nuance.

AI-powered search with natural language processing and personalized content recommendations are more challenging to implement, as they require a deep understanding of user context, preferences, and behavior. These features may also need continuous optimization and fine-tuning to ensure optimal performance and user satisfaction.

Conclusion:

To conclude this part, we can classify the features based on the addition of the two scores defined above.

Ranking of the features based on the sum of their scores:

1. Language translation (16)
2. Grammar and orthography correction (16)
3. Automatic content tagging for organization (15)
4. AI-powered search with natural language processing (14)
5. Speech-to-text (audio transcription) (14)
6. Automatic article writing from a title (13)