Decoding CX -Customer Experience — In the age of Gen AI

[Note — These are all my personal opinions,this has no bearing to any company or industry]

CX is a broad term used by industry people to group several use cases all related to delivering an end-to-end seamless customer experience from prospecting to onboarding to customer support and more. A lot of technology is needed to solve these use cases and several companies have been created to solve these use cases from simple customer communications to contact centers to CRMs to name a few. In this blog, I am mainly discussing the Contact Center side of CX.

Contact Centers have evolved a lot both in terms of technology and in terms of the business model. From on-prem to the cloud to SaaS, everything has changed in the last decade, and it's rapidly evolving in every aspect. Voice technology has seen a lot of evolution from PSTN ,VOIP, WebRTC to mention a few. The number of channels has proliferated from Voice to Omnichannel such as SMS, Text, RCS, and Social. The Contact Center Software providers and their customers have steadily embraced all these changes. The Contact Center industry is again ready for a step change with AI and Gen AI and is rapidly embracing this change in every aspect.

Let us break down the Contact Center software space into three component phases and analyze how AI can transform each of these components.

Phase 1 > Pre-Call: This is before a customer call or interaction reaches any human agent. AI has a lot of use cases to solve in this space. Example are :

IVAs (Intelligent Virtual Agents) — They can interact and solve a customer’s problems just like a human agent. IVAs are powered and designed by interaction flows (Example Five9 Inference Studio, Google Dialog Flow, AWS Connect Flows). Gen AI is even at a stage where the Conversation Flow designers will themselves be AI and AI will be designing Interaction flows which in turn will be powered by AI.

Bots — These bots can range from SMS, Chat, and WhatsApp bots powered by Gen AI models, with a human-like interaction

Synthetic Voice — AI can produce a human-like voice that will sound and feel like a real human, almost passing the turning test.

NLU — Natural Language understanding algorithms and models powering the IVAs

Sentiment Analysis — Gen AI models understanding and predicting the mood and customer sentiment and deflecting the conversation accordingly.

With the improvements in GenAI in a few years we can expect a lot of mundane interactions to be 100% served by these AI agents before they could even reach the human agent. This would mean 100X productivity gains by the Customer Support and Sales teams and great experiences for end customers.

Phase 2 > Live Call: This is when the AI Agent was not able to solve the customer’s problem and the customer wants to interact with a human agent. The ways AI can influence this space are

Agent Assist — AI-powered Agent Assist software can provide intelligent hints, assistance, and knowledge base article searches in real-time helping the agent to help the customer in real-time this means every agent or sales rep will be a star agent.

Live Speech-to-text transcriptions and translations — AI can do live transcriptions and translations and also record the data into your specific CRM of choice for future analysis. This can also be used by the Supervisor to analyze a conversation.

Live Sentiment Analysis — AI can understand the mood and sentiment of the customer and the agent and provide hints to steer things in the right direction.

Call Summarization, Disposition, and Wrapup — AI can eliminate call wrapup time to almost Zero by auto data entry, summarization and compliance checks to name a few. This means agents are ready quickly to handle the next customer which means less customer queues and happier customers.

Phase 3 > Post-Call: This is the phase that will use all the data collected in the first two phases and create the adequate data lake and models for predictive analytics, Workflow optimization, and Quality Management to name a few all of these could be done by AI and work in this phase will make phase 1 and phase 2 even better and more intelligent.

My Hypothesis: Contact Center Industry is one of the most exciting industries at this point to really make AI work. There will be hundreds of startups created solving granular components in each of these phases. Most of these startups would be at some point acquired by the larger Communications and contact center players, the reason being, from a customers standpoint all these components need to work together as a single offering in tandem to provide full value to customers and piecemeal solutions cannot really solve a customer’s problem end to end. Only Full-scale solution providers will be able to address these problems holistically. However, this will be one of the most exciting spaces for everyone — Established players, startups, entrepreneurs, and investors.

Keep building great stuff.