Blog Post

Khoros Insights Blog
5 MIN READ

Exploring AI Use Cases in Khoros Communities

LarryI's avatar
LarryI
Khoros Expert
2 years ago

Welcome to the Khoros view on our future with AI!

It really is an exciting time for online communities in many ways, but we are just at the beginning of exploring how this technology will impact how we work in communities.

At Khoros, we have been hard at work on Aurora, our next generation of online communities.
While we have focused on our new technological foundation, natural language processing, and large language models have come to dominate the conversation in tech. Naturally, this has led to plenty of exciting discussions about the future of online communities. With Aurora, we are focused heavily on developing robust API’s, decoupling the UI from the backend of our platform, and building modern app experiences that will enable us to quickly tap into future trends in technology, whether it’s the metaverse or generative AI.

What is Khoros working on now to further innovation in this space?
There is much to explore with generative AI as it is still very new, but we will experiment and are planning on incorporating new technologies into our products.  

“One of the biggest initiatives we are working on is an orchestration layer for customer engagement that sits between these services and direct engagement with users offering the ability to control the use and workflows of that engagement. Whether it's easily swapping or testing new services, or putting the technology in front of humans to experiment and refine, we believe this will be a critical component of enabling and controlling the many use cases that these services make available.” - Nick Hill, VP, Product - Communities at Khoros

There are some focus areas for us in product development with the application of generative and conversational AI.
The most notable is the relationship between community-generated content and self-service / discovery. Five out of the top 10 sites on the internet today are communities of peer content. Communities rely heavily on SEO to index this content and on searches to drive traffic to the community. Generative AI offers a new approach - but it's still heavily reliant on the continuous sharing of user generated content. With the incredible scale of 18M new posts per year and hundreds of millions of solution views across all Khoros communities, the fundamental need for experts to share, and the need for others to find is unchanged. However, the approach of crawling and links will shift to summarization and intelligence responses. Some exciting new capabilities will be to build on your query, or to generate questions where knowledge is insufficient. 

Another opportunity we have is exploring the improvements for content moderation.
Generative AI tools can help us better caption and classify images and text, which reduces the need for manual moderation and helps improve moderator mental health. This addresses spam and inappropriate content in new ways and adds much more to the moderator's set of tools. However, we’ve also seen ChatGPT, among other Generative AI tools, used to increase human-like content spam in communities for nefarious purposes. Consequently, focus on these tools and implementation of guardrails around these tools is paramount to the success of community moderation going forward.

Content is key in every community, and the focus should absolutely remain on human-created content.
When it comes to the content generation space, brand communities are often specialized knowledge sources and their existing content can help fuel creation of new content such as guides on how to use a product, marketing materials about use cases, value stories, etc. The usage of these tools depends on the expertise and validation of humans. These tools can be leveraged for suggestions on how to improve the content, tag your content for search optimization, and help guide you to follow the community guidelines. Generative AI is better suited to aiding humans in the creation of valuable content versus being used solely to generate the content—especially given how prone to errors it can be at this point. Personalization on the user front is another very important use case to consider as we work toward leveraging AI tools to surface content relevant to the users’ interests.

Beyond the end-user impact, we’re also exploring the potential impacts in the community management front.
Managing relationships with your community is key to your success. We see possibilities to apply AI to the user onboarding process as well as tracking and bolstering engagement with your most active, loyal users. Imagine being able to derive analytics, insights, and trends all relevant to your query. This knowledge would give you more power in determining your community strategy and help direct you toward your next steps. Understanding your users’ sentiments and challenges, along with surfacing the most impactful tasks you can focus on, will be key to supporting community management in the future.

Additionally, we are exploring support use cases with agents. Agent efficiency is a key metric for any enterprise.
Therefore, we are seeking to combine knowledge in the community with previous agent responses to better assist agents in resolving cases quickly and accurately. We are investigating suggested responses, incorporating community knowledge to better assist agents, and exploring the ability to summarize a case and transform it into reusable knowledge that could be posted back to the community for self-service. For example, in a marketing use case, we are investigating how to generate content of different lengths and keywords to publish across channels, including recommendations on how to tag and
when to publish.

We’ve been working on automation, text classification and NLP for a long time and they are key parts of many of our solutions.
With that existing framework, there is plenty of opportunity at Khoros to continue to integrate AI technologies in our products, and we are very excited to work to include these tools in Communities. 
The importance of having a platform, generating content, and strong community experiences focused on engagement with extensible platforms like analytics, metrics API, and firehose all provide us with the right foundation to ensure brand communities and user-generated content continue to play a leading role in the future customer experience.  

 

What are you most looking forward to seeing happen in the community platform? What are some challenges you are facing today?
Let us know your thoughts in the comments!

 

Join the discussion on these great topics in the Atlas community!

Check out our podcast episode covering The Future of Conversational AI, with Khoros’ Gijs Van De Nieuwegiessen.

Updated 2 years ago
Version 4.0
  • Thanks LarryI ! One of the key capabilities of AI is to see trends from large amounts of data, which is very hard for us humans to do, would be great to explore and build this out, so that we get accurate actions we can take based upon these trends, a "predictive AI." Another is better and more powerful sentiment analysis and as you tease above, possible proactive action from an AI. So many possibilities...

  • Hopefully we'll see some improvements in accessibility text automatic generation as a byproduct of this.