PepsiCo Uses AI Twins, SAP’s AI Tax Overhaul, & Travelers’ Claims Shift?

In partnership with

AI HUSTLE | February 3, 2026

Welcome back to AI Hustle, the newsletter that turns AI theory into business reality. This week we’re looking at how AI is moving from the chat window to the factory floor and the core of our operations. It’s easy to get distracted by flashy demos, but the real money is being made by integrating AI into the unsexy, essential workflows that power a business. Today, we’ll build a system to predict customer churn before it happens and then dive into how giants like PepsiCo and Travelers are using AI to reshape their physical and digital infrastructure.

Start learning AI in 2026

Everyone talks about AI, but no one has the time to learn it. So, we found the easiest way to learn AI in as little time as possible: The Rundown AI.

It's a free AI newsletter that keeps you up-to-date on the latest AI news, and teaches you how to apply it in just 5 minutes a day.

Plus, complete the quiz after signing up and they’ll recommend the best AI tools, guides, and courses — tailored to your needs.

Sign up to start learning.

The Hustle: The Proactive Churn Shield

The Goal: Automatically identify high-value, at-risk clients before they cancel their subscription, giving you a chance to save them.

The Tools:

* Product Analytics Tool (e.g., Mixpanel, Amplitude)

* Helpdesk Software (e.g., Zendesk, Intercom)

* Automation Platform (e.g., Zapier, Make)

* Slack

* A Large Language Model (e.g., GPT-4 via API)

Step 1: Gather the Signals (The Input)

Your business is constantly generating data that signals customer health. This hustle combines two critical streams. First, use your product analytics tool to track user login frequency. This is your behavioral signal. Second, use your helpdesk software to track all inbound support tickets. This is your sentiment signal. The goal is to have a live feed of how often key clients are using your product and what they're saying when they reach out for help.

Step 2: Define the Red Flags (The Trigger)

This is where you set the rules for your early-warning system. In your automation platform (like Zapier), create a workflow that constantly monitors your inputs. The trigger is a combination of conditions. For example: "IF a customer on a 'Pro' plan or higher (a high-value client) has NOT logged in for 10 days, OR IF that same customer has submitted two or more support tickets in the last 30 days that are flagged as 'Frustrated'." This combines a lack of engagement with active frustration—a classic churn precursor.

Step 3: Let AI Connect the Dots (The AI/Logic)

When a trigger condition is met, the automation kicks in. First, it uses an AI model with sentiment analysis capabilities to scan the text of the recent support tickets. It looks for keywords, tone, and phrases that indicate frustration, anger, or confusion, and flags the ticket accordingly. If the "two frustrated tickets" rule is met, the workflow proceeds. Then, it feeds the customer's name, their issue history, and their lack of recent logins to a generative AI model (like GPT-4) with a specific prompt: "Draft a friendly, non-intrusive 'personal check-in' email from an Account Manager to [Customer Name] who seems to be having trouble. Acknowledge their recent issues and offer a quick call to help."

Step 4: Alert and Equip Your Team (The Output)

The workflow doesn't act on its own—it empowers your team. It performs two final actions. First, it sends an immediate, detailed notification to a dedicated Slack channel (e.g., #churn-alerts). The message should summarize the situation: "ALERT: High-value client [Customer Name] is at risk. No login for 10 days and 2 frustrated support tickets. Proactive outreach recommended." Second, it creates a draft email in your Account Manager's Gmail or Outlook, using the AI-generated text from Step 3. All the manager has to do is review, personalize it slightly, and hit send.

Why This Hustle Works:

* It's Proactive, Not Reactive: Most companies wait for the cancellation email. This system puts you ahead of the problem, turning customer success from a reactive fire-fighting role into a proactive, revenue-protecting one.

* Focuses on High-Value Accounts: By filtering for your most valuable customers, you ensure your team spends its time on the accounts that have the biggest impact on your Monthly Recurring Revenue (MRR).

🚀 The AI Pulse: 3 Signals to Watch This Week

Pepsi's AI Bet Isn't on Chatbots, It's on Factories

PepsiCo is using AI and "digital twins"—virtual models of its physical factories—to simulate and optimize production line layouts and operational changes. Instead of risky, time-consuming physical trials, they can test thousands of scenarios virtually to find the most efficient configuration. This approach compresses planning cycles from months to weeks and de-risks expensive capital investments in its core operations.

The Hustle Take: You don't need a multi-billion dollar factory to use this playbook. Any business with complex physical operations—from e-commerce warehouse layouts to retail store flows—can use simulation to test changes before spending real money. The AI opportunity isn't just in knowledge work; it's in using models to de-risk costly decisions in the physical world.

The UK's Tax Agency is Rebuilding for AI from the Ground Up

The UK's tax authority, HMRC, is undertaking a massive infrastructure overhaul with SAP, moving its core systems to a secure, AI-ready sovereign cloud. Instead of just adding an "AI layer" on top of legacy tech, they are rebuilding the foundation to ensure their data is unified, accessible, and secure. This allows them to build machine learning directly into core processes like tax administration and taxpayer communication.

The Hustle Take: AI is only as good as the data it runs on. This is a critical reminder that true AI transformation often starts with "boring" infrastructure work. Before you spend big on the latest AI tool, ask if your data is clean and accessible. HMRC's move proves you need to get your data house in order before AI can deliver real value, especially in regulated industries.

How Travelers Insurance Used AI to Cut Call Center Staff by a Third

Insurance giant Travelers has embedded AI across its business, leading to massive efficiency gains. By automating claims processing, using AI to analyze risk for underwriters, and deploying a natural language voice agent for initial customer calls, the company has reduced its claims call center headcount by a third. Over 50% of claims are now eligible for straight-through automated processing, dramatically reducing handling times and operational costs.

The Hustle Take: The biggest AI returns come from augmenting specific, high-volume workflows. Travelers didn't just give everyone a general-purpose chatbot; they targeted the repetitive, data-intensive processes at the core of their business—claims and underwriting. Map your most critical operational workflows, find the biggest bottleneck, and focus your AI efforts there. That’s where you'll find your ROI.