October | 2021
The ongoing pandemic has forced organizations of all kinds to transform their customer experience. Recent research by Frost & Sullivan shows that 83% of respondents say improving customer experience is the most important strategic business goal to help retain existing customers and gain new ones. One of the ways organizations are looking to bolster the customer experience is by improving customer interaction channels with chatbots. In fact, 80% of businesses expect to use some sort of chatbot automation by the end of 2021, and the chatbot market is projected to grow to $9.4 billion by 2024 with a CAGR of 29.7%.
Research around systems with good natural language understanding (NLU) capabilities has accelerated, helping conversational agents improve the way they understand human language and conduct more human-like conversations without the need for manual intervention. This dramatic improvement in chatbot technology has enabled many organizations to embrace chatbots as their first line of response.
Chatbots Go Mainstream
Chatbot uses have proliferated recently due to the drastic decrease in available on-site workforce and the inability of most contact centers to shift operations to a work-from-home model. This wide chasm in the way the organizations respond to customers has opened a new avenue for automation.
While the contact center use-case is one where organizations had to implement a chatbot application, chatbots capabilities can improve the user experience in many scenarios. A few examples include improving the experience of an in-vehicle infotainment system, helping customers book airline tickets, or helping assemble home furniture.
The early days of conversational applications involved conversations with a single individual based on simple textual patterns. Today, the technology has scaled to simultaneous conversations with potentially millions of users. The elasticity of the cloud has made this possible.
Scalability was one of the major drivers for organizations to first move to the cloud. With the evolution of modern conversational agents, building a true chatbot application has become much easier. On-tap availability of cloud resources with a click of a button means that scalability and durability are now a given. It is simply left to teams to define the threshold limit rather than worry about the resource provisioning.
Given the elasticity of the cloud and the shared security model of the applications, cloud has become the testbed for growth. The shared security model emphasizes that the onus on security is a two-way street. Essentially all participants – the consumer of cloud services and the cloud service provider – have a responsibility to secure the cloud.
The evolution of conversational applications from a simple app providing basic answers to user queries to the underlying use of AI/ML has been phenomenal. The rapid development of the AI technology in chatbot applications has made sure user queries are more contextual and provide intelligent responses to the user. The important components that constitute a fully functional conversational application include:
The dialog system is the heart of the conversational application. All user utterances or inputs are first fed into the dialog system. These are mapped to the respective intent in the conversational agent, which quickly detects the correct intent and recommends or takes a suitable action.
The backend fulfillment is where the customization of responses is handled. The backend fulfillment consists of lightweight code that accesses the database. Based on user utterance, specific predetermined responses are fetched. The database serves as the single source of truth that can be used to store the customized responses or to provide inputs based on the conversation. These components make the base framework for a fully functional conversational application, but other services are needed (like cloud storage and big query).
Security is of paramount importance when it comes to the cloud. With the recent breaches of Solar Winds and the Colonial Pipeline hack, securing the cloud is the topmost priority for all organizations with a footprint in the cloud. One of the areas that needs to be addressed is the hi-jack of the conversations which might lead to data leakage or even prove life-threatening. Address security concerns head-on by securing data and include government regulations on how the customers data is handled and used.
The Egregious 11 cloud computing threats report from the Cloud Security Alliance provides a framework of the threats to consider while building highly resilient and scalable global applications on the cloud. While these data threats are not discussed in detail, here is a list for reference:
The threats above apply to the cloud computing environment. Serverless workloads have their own unique operational and security challenges. This list of best practices by the Cloud Security Alliance can be used to design a secure and resilient serverless architecture:
These are a few of the threats that should be addressed while designing a secure and resilient application in the cloud. The Security Command Center (SCC) provided by Google Cloud is a good starting place to detect these threats and take corrective or preventive action. SCC provides runtime security for secure production workloads without having to worry about any data loss or malware attacks at runtime. In addition, the Open Web Application Security Project (OWASP) provides a list of the top 10 vulnerabilities and therefore risks to consider while designing secure web applications.
Conversational applications are finding a wide range of uses across a variety of domains, from automating the customer engagement to an organization’s first line of defense troubleshooting customer’s issues. With the rapid advancement of conversational AI and by extension NLU, conversational platforms like Google’s Dialogflow can be used to build efficient highly scalable real-time conversational applications that have a multitude of applications in every domain from automotive, healthcare, telecom, banking, and retail.
Gartner stated in 2019 that “by 2025, customer service organizations that embed AI in their multichannel customer engagement platform will elevate operational efficiency by 25%.” The pandemic expedited this move. Organizations everywhere are scrambling to figure out how they can have meaningful engagement with their customers as traditional channels are no longer relevant.
From banking to telecom, organizations are transforming the way they engage with their customers. Banks have deployed virtual assistants on their web and mobile channels. These assistants can be activated by Google Assistant or other popular voice assistants. This is transforming how customers interact with companies.
Telecom operators are deploying conversational assistants in their set-top boxes to help customers interact with the devices in a more natural way. Now, customers can not only voice control the TV but can control home appliances using the voice-based assistant.
Automotive giants are looking to improve the way users interact with cars. By deploying the smart voice assistant to augment the in-vehicle infotainment system, carmakers everywhere are looking to gain more insights on how users interact with their cars and use these metrics to improve the overall experience. Conversational assistants can help drivers by providing suggestions on intelligent predictive maintenance that ultimately decrease overall service costs.
Conversational analytics gathered by reviewing the way users interact with voice assistants can improve the customer experience and provide valuable insights about the end-user journey. Some of these real-life journeys with chatbot applications could be overlooked, as it is difficult to predetermine every customer need. Insights gathered in this way can then be used to visualize the most ubiquitous journeys of the users, further automating the end-user experience.
The applications for the conversational assistant are endless. If deployed correctly, these assistants will go a long way in satiating the organization’s need to automate customer engagement while collecting valuable metrics and insights that will give customers a true 360-degree experience.
About the Author
Sriharsha Makkuva
Solution Architect, Cloud Product & Platform Engineering Practice, Wipro
Sriharsha has more than a decade of IT experience and is currently a solution architect building highly scalable and resilient applications on the cloud. He has worked across multiple enterprise-wide digital transformation projects in the retail, banking and automotive domains. Sriharsha has published papers on digital transformation and digitalization in various national forums.
This white paper explores the drawbacks in existing QFR technique and also proposes solution which can improve QFR success rate.
How a comprehensive framework for efficient network testing can help operators deploy LTE networks faster
This paper explores the possibility of eMBMS becoming the standard technology platform for transmission of multimedia content over cellular networks. It details out the advantages of adopting eMBMS and the various scenarios in which it can play a game-changing role
