In the digital transformation of any company, AI/ ML is one of the crucial tools that can be leveraged. At Digité, we are keenly aware of the need for AI/ ML to assist and enhance capabilities our solutions provide to our customers in all possible manner.
We look at AI in two distinct categories. There’s Ubiquitous AI which includes applications that solve short, highly contextual objectives using simple data science pipelines. These applications serve as a way to improve the quality of user’s experience of any product they use.
Then there’s Transformative AI which includes applications that involve more complex data science pipelines that are used to bring change in a fundamental way in the user’s method or application, and not just merely improve it. Very often, Transformative AI products might be standalone tools that take their data from other – perhaps multiple – applications – and can be used to predict and improve business outcomes in a wide variety of businesses and applications.
In a myriad of ubiquitous AI applications, text similarity is one such use case that can be applied to various facets of work. If you are part of a Support team or a Support function, how often have you wished that you could fill fields with readymade ticket descriptions generated from similar tickets that were solved previously?
Specifically, for high contact functions like Support whose performance is measured by quick turnaround, customer satisfaction and team knowledge, injecting AI-capabilities in small doses can have a dramatic on the users’ ability to resolve business issues or complete business workflows much more speedily and with greater accuracy.
In this post, we briefly describe the high impact of a Similarity Engine in the context of the Support function. A typical support team may address 100s of issues (tickets) in a day or a week. In the course of doing so, it encounters the following issues:
Similar or the same issues are raised or logged as support tickets repeatedly by multiple customers.
Worldwide research has found that more than 13% of all incidents are repeat incidents, resulting from the inability to accurately determine the root cause of problems. Just like with repeat incidents, organizations are struggling with problems being logged multiple times and handled in parallel. More than 17% of the total ticket volume are duplicates. (on top of the 13% of repeats which are tickets reporting the same problem but different wordings, as shared in this report sponsored by Splunk)
In high-impact projects, customer service representatives have to often depend on documentation or other seniors for knowledge and to ensure that ultimately, they provide high quality of resolutions!
The average large organization can expect 852 duplicate or repeat incidents in a month! This results in wasted IT resources and subsequent costs as well as reduced customer satisfaction. (Taken from the same Splunk Report quoted above)
This is where apps like Ditto help. Ditto is a semantic search engine that automatically identifies tickets that are similar to the current open ticket that has been selected.
Semantic similarity is a concept in the Natural Language Processing (NLP) domain that is used to find sentences or words with similar meaning or context. Semantic similarity has many important applications in NLP such as information retrieval, text summarization, question answering, relevance feedback and text classification.
Ditto uses neural networks to rapidly build models in memory and present results. Semantic similarity is sought by analyzing historical tickets, and then trying to see if the new ticket is similar in context to any of the historical ones. This helps find existing solutions that have solved previous tickets that can also be applied to the new tickets.
Ditto has been specifically created for Zendesk users and works directly within the Zendesk application interface by providing relevant information with accuracy and immediacy.
Security! Security! Security!
Ditto ensures complete configurability to determine who sees what! This ensures that there is isolation of knowledge within specific teams as per the need of the user. Different models are built in such cases and hosted separately to ensure security.
Location! Location! Location!
We ensure that all information from the AI engine is readily available in the location our users need it the most. So, there is no separate interface that one needs to go for relevant information.
So, if you are a Zendesk user, we hope that you leverage this awesome AI app that we have created for you on the Zendesk Marketplace!! You can give Ditto a Try here: https://www.zendesk.com/apps/support/ditto/
P.S: Coming soon!
Auto-fill – It can prove to be tedious to fill forms often. Auto-fill aims to relieve us of this pain. When you have to fill lots of fields, auto-fill helps fill these fields automatically.
Sentiment Analysis – While similarity compares new tickets with historical ones, sentiment analysis concentrates on figuring out the user’s intent by analyzing chat content related to the user’s ticket. This helps understand the purpose behind the user’s ticket creation.