ET Bureau: The dissemination of knowledge through the enterprise is key to its best utilization. However, in most companies, there are clear hurdles to this process. What challenges have you seen in this process?
Historically, the dissemination of knowledge represents a problem for an enterprise. Not just the dissemination but retention of knowledge as well. It has been a passive approach. By that, I mean, there is no closed loop between creation, usage, and refinement of knowledge. The result is that people get frustrated because they are not getting the information they need, and so will stop using the tools that have access to knowledge. What is required is an active approach. The problems companies face can be summarized in three key areas:
How to solve the challenges in adopting Artificial Intelligence
First, there is a knowledge gap. Information is either missing, incomplete, or there is too much information that overlaps or conflicts. This gap refers to the fact that the knowledge seeker cannot access what they need when they need it. The result is slow or incomplete responses to queries. Or even worse- receiving conflicting information.
Second, the knowledge seeker differs by individual. How they ask questions, the keywords or phrases that are used can vary widely. This results in an attempt to solve for the masses, which is at the lowest common denominator and means that basic information is being served up that may or may not apply to the request if the individual is an “expert”.The other side of the coin being the information may be too dense (lots of documents or an intense level of detail) that make it difficult to serve up the answers in a way that is consumable for a more novice user. Existing tools are neither complete nor adaptive enough to meet the needs of knowledge seekers.
Artificial Intelligence is Becoming the Future of Investment Platforms
Finally, the third challenge is actually connecting the dots. Companies need to serve up the knowledge to the requestor in a way that is efficient and relevant. It is one thing to acknowledge you have a gap, and it is another to understand how to close it. Besides, a process of continuous improvement is needed to update and serve up what best meets the needs of the knowledge seeker. Machine learning plays a vital role in automating the orchestration of this.
ET Bureau: How is AI changing support interactions? Is it adding more value to the process of support? Or do you see teams finding it a challenge to understand clearly?
At Serviceaide, we are committed to leveraging AI to digitize service management for maximum impact, satisfaction, and value. Over the last 12 months, we see an acceleration in interest and opportunity where our sales managers have active projects to leverage AI for improved support and service interactions. Interest in Luma virtual agents is increasing as the market matures and recognition that this newer technology can solve the existing challenges I mentioned.
Specifically, the need to map and understand patterns of requests to the available information. By normalizing a vocabulary to support semantic analysis, this information can then be contained in a knowledge graph for contextual relevance. The combination of these three is critical to make the knowledge accessible and available to all requesters at every level.
Artificial Intelligence Infused with Big Data Creating a Tech-driven World
The value of an intelligent virtual agent with automation and knowledge is very tangible. As companies improve request responses, both providing the right information as well as speeding up the process, they also optimize through directed assisted learning. The method of service improvement is not an onetime project but an ongoing service and knowledge initiative. The savings are qualitative and quantitative, including increases in requester satisfaction and productivity gains.
We absolutely see an increasing understanding of the challenge both in actively engaged projects as well as new conversations, why this problem is essential to solving, the desire to address the situation, and more adoption across medium to larger enterprises.
ET Bureau: Very often, in an enterprise, there is a clear gap between users and developers of technology. How can AI help to bridge that gap?
I used to run the global support centers for a large software vendor. Self-help and self- education for users at that time were limited. There was either a user manual that walks end-users thru step by step scenarios or “online” help. The online support was simplistic, where “how to use” content was delivered in the context of the GUI screen to the end-user. In addition, there may have been education courses/workshops to teach the proper/prevalent use cases. These are all generally very heavy in implementation and usage, taking time from the developers of the technology to cast themselves in the mindset of the end-user (to develop content) and also time-consuming for the end-user to either read or attend sessions. It was not always productive or valuable.
Fast forward to 2020, where AI is transforming the world– including service and support management, where AI can deliver an intelligent mediation element. By leveraging the knowledge mentioned previously, AI can provide a common pattern context in an evolving vocabulary, intent, and content framework. This allows the disambiguation of available content between the service provider and the service consumer to get the right answer at the right time.
ET Bureau: What value do you see AI adding to the process of service automation? What could be the biggest challenges to watch out for in the adaption of automated services?
The simple value of leveraging AI with service automation is getting the right answer or action at the right time. There are many areas in our daily support and service interactions where we can speed up responses and automate them. From simple actions like password resets to more complicated and more enduring interactions in a business like an HR request for benefits information or time off.
I believe a key challenge to the adaption of more automation will be around “trust”. This applies to the organization implementing the technology as well as from the end-user. It is important to trust the process and trust that the relevant knowledge is served up. An auditable virtual agent should allow automated services to be reviewed and validated periodically. An administrator can review a dashboard to view how things are presented, what was done, and if it was successful, etc. As we increase expectations around response times and service excellence, just as with human labor, trust is an essential threshold for digital labor solutions to excel.