AI in Business - Week 9

AI Concierge: Redefining Customer Service

•From "Press 1 for English" to empathetic Virtual Assistants

•The end of "Hold Music" frustration

•The transition from robots to partners

Customer Service

Reactive: Fixing the broken or after it breaks

The assistance gives after an interaction fails

What an example of customer service?

Calling an agent because a flight was cancelled

Customer Service (CX)

Proactive: Total Interaction

Every interaction you have with a brand - the app, the seat comfort, the reminders

The total sum of all touchpoints a customer has a brand

What an example of customer experience?

Easy navigation, proactive reminders, and frictonless usage

What does good customer experience prevents?

The need for customer service

The Ultimate Business Goal

Engineer friction out of the CX loop entirely to eliminate the reactive cost of CS

Great CX prevents the need for customer service in the first place

What is the ultimate win for a business

If CX is perfect — meaning the app is easy and flight is on time —- the customer never has to call service at all

Which is a subset of each other?

Customer service is a subset of customer experience, ensuring customers never find the need to call customer service

How Do We Measure Happiness

•NPS (Net Promoter Score):

•Would you recommend us?

•CSAT (Customer Satisfaction):

•How was this specific chat?

•FCR (First Contact Resolution):

•Solving it the first time

•SLA (Service Level Agreement):

•The company’s promise (e.g., "reply in 24 hours")

Which measurement is big for AI?

FCR - every time an AI solves a problem on the first try, the company saves massive amounts of money, and the customer stays happy

Net Promoter Score - NPS

How likely are you to recommend us?

Separates Promoters from Detractors

Measures customer loyalty based on promoters (score 9–10). Example: 78

Customer Satisfaction - CSAT

Measures a single, isolated interaction

How satisfied were you with this specific touchpoint?

First Contact Resolution - FCR

Did we fix it on the first call

High FCR directly lowers business costs by eliminating 3x callbacks

Service Level Agreement - SLA

The operational promise a company makes

Ex: Guaranteeing an email reply within 24 hrs

What plays a huge factor with these metrics?

Churn where customer churn is a key area where ML and AI is commonly used

Proactive Customer Experience (CX)

Proactive Customer Experience eliminates the need for reactive Customer Service

Customer Experience (CX) vs Customer Service

CX is proactive across the entire journey, while Customer Service is reactive damage control during support.

What is the business goal of proactive customer experience

Prevents the SLA (Service Level Agreement) from ever being tested

The Evolution of Support: From FAQs to Rag

•Past: Rule-based bots that just say "I didn't understand that. Here is a link to our FAQ".

•Present: Generative AI Agents that are context-aware and connected to company data (RAG).

•Real World Case: Klarna’s AI did the work of 700 human agents in its first month with 25% fewer repeat inquiries

Ex: Klarna’s Support

It’s not just about speed; it's about being "context-aware"—meaning the AI knows who you are and what you bought before you even say hello

Generative AI in Customer Support

Generative AI is obliterating traditional support bottlenecks.

The Scale of AI Support

AI can handle millions of customer interactions simultaneously, replacing large human support teams.

Predictive Analytics in CX

Seals the leaks before the customer ever complains through invisible signals, AI risk engine, and action pipeline

Invisible Signals

Customer behaviors such as checking contract end date on app, dropping 3 calls in one week, and stop opening marketing emails.

AI Risk engine

System correlates disparate behaviors and flag users for imminent departure

Action Pipeline

System automatically triggers a “20% off your next bill” SMS offer before the user ever initiates a cancellation call

Proactive Rentention

What is the goal of predictive analytics?

Solve the problem before the customer even knows they are unhappy

What does predictive analytics do?

Flag churn risks before the customer ever complains

What are the steps of predictive analytics?

Step 1. Invisible Signals

Step 2. AI Predictive Flag

Step 3. Automated Action

Social Sentiment at Scale

AI instantly scrapes 10,000 messy reviews and synthesizes actionable truth: 70% of negative sentiment in the last 48 hours is tied to the new lid design leaking

Sentiment Analysis at Scale

Social Listening

Summarization

Hyper-Loyalty

Social Listening

AI scrapes TikTok, X, and Yelp to gauge the “mood” around a brand

Summarization

Analyzing 10,000 reviews to find one specific issue (e.g., “The new lid design leaks”)

Hyper-Loyalty

Starbucks creates unique challenges for you based on your specific past order

What kind of factor does Sentiment play?

The core of customer AI and customer experience

What does social listening do?

Transform unstructured noise into hyper-loyalty at scale

What is an example of Sentiment Analysis at Scale

10,000 Unstructured App Reviews:
AI analyzes large volumes of unstructured customer feedback.

Engine / AI Business Logic (Sentiment Summary):
Pattern detected: 70% of negative reviews in the last 48 hours mention the new lid design leaking.

Surface / Consumer Experience (Hyper-Loyalty Action):
AI generates a unique “Bonus Star Challenge” based on that individual’s specific order history.

How Streaming Actually Makes Money

SVOD

AVOD

You

SVOD