Why AI Applications Could Be the Next Big Opportunity: Lessons from Beerud Sheth

Why AI Applications Could Be the Next Big Opportunity: Lessons from Beerud Sheth

I began my conversation with Beerud Sheth on Dilse Omni Talks with a simple question: many of us know Gupshup through an Indian lens, but how global is the business today?

Beerud is the Co-Founder and CEO of Gupshup, a global conversational AI platform for marketing, commerce, and customer support. An IIT Bombay and MIT alumnus, he previously co-founded Elance, which later became Upwork. At Gupshup, he has helped build an India-focused messaging business into a global conversational AI company.

What followed became a masterclass in three connected ideas: how to build locally while scaling globally, where the next AI application opportunities lie, and why customer conversations cannot remain confined to individual channels.

My biggest learning was this: the opportunity is not simply to put AI into a business. It is to make AI useful within a specific customer context.

Table of Contents

  • 1. Global Scale Requires a Multilocal Mindset

  • 2. The Next AI Opportunity Is in Applications

  • 3. The Same Messaging Channel Can Behave Differently Across Markets

  • 4. Voice Notes Are a Product Requirement, Not a Small Detail

  • 5. Omnichannel Means Moving With the Task

  • My Biggest Takeaway: Build for Context, Not Just Capability

1. Global Scale Requires a Multilocal Mindset

Beerud shared that, at the time of our conversation, more than half of Gupshup’s revenue came from outside India. The company served over 50,000 customers across more than 100 countries, with people based in over 22 countries.

That is a significant shift from the largely India-based operation he described from four or five years earlier. Expansion across WhatsApp-heavy markets, including Latin America, the Middle East, and Southeast Asia, had changed the business.

But the numbers were not the most interesting part. The operating philosophy was.

“Not as a global or international company, but a multilocal company.”

Beerud Sheth

His explanation was direct. In Brazil, you have to feel like a Brazilian company. In Mexico, you have to feel like a Mexican company. Customers judge you against the alternatives available in their market, not against the difficulty of your international expansion.

A global perspective can be valuable, but it cannot substitute for local understanding.

The multilocal operating framework

I would translate Beerud’s approach into three connected building blocks:

  • Local understanding: Build teams that understand the market and its customer expectations.

  • Local authority: Give those teams the ability to get things done, rather than treating them only as sales outposts.

  • Shared learning: Bring customer and product feedback from each market back into the wider organization.

Global perspective + local experience + empowered teams = a multilocal business.

For me, this reframes expansion. It is not just a distribution exercise. It changes hiring, organizational culture, product feedback, and everyday decision-making. A company cannot promise local relevance while keeping every meaningful decision disconnected from local customers.

2. The Next AI Opportunity Is in Applications

Our conversation then moved from Gupshup’s expansion to the broader AI opportunity.

Beerud pointed out that much of the attention and investment has gone into chips, data centers, and frontier models. His argument was that the next major opportunity lies in applying those capabilities to real business problems.

“The next big opportunity is AI applications.”

Beerud Sheth

This distinction matters. A powerful model provides capability. An application has to turn that capability into something useful for a particular business and customer.

Beerud also sees the geographic opportunity widening. While the US and Western markets have traditionally been more web-centric, he noted that conversational AI, wider RCS availability, and improving voice AI models are creating new possibilities there too.

So the opportunity is not limited to markets where WhatsApp already dominates. The ways people interact with businesses are evolving across regions.

Beerud’s opportunity cube: Geography × Industry × Use Case

One of the clearest frameworks he offered was a three-dimensional cube. Each axis represents a different choice:

  • Geography: Where is the customer?

  • Industry: What business context are we serving?

  • Use case: What specific task needs to be completed?

GEOGRAPHY × INDUSTRY × USE CASE → A specialized AI application opportunity

Every intersection creates a distinct set of requirements. Beerud illustrated this with three examples:

  1. Dubai → Real estate → Lead generation: An AI-powered customer experience specialized for that market and domain.

  2. India → Ecommerce → Order tracking: An experience shaped by local commerce workflows and payment methods such as UPI.

  3. Brazil → Fintech → Credit card sales: An application that supports local requirements and payment methods such as Pix.

What I liked about this framework is that it replaces a vague question, “What can we build with AI?”, with a much sharper one: Which customer task, in which industry, in which market, can we solve well?

The model may be broadly capable. The application still needs to be specific. That is where local knowledge and industry understanding become essential.

3. The Same Messaging Channel Can Behave Differently Across Markets

I wanted to understand what changes when a business serves Latin America, India, the Middle East, and other regions. If the underlying messaging platforms are similar, how different can customer engagement really be?

Beerud’s answer showed why copying a successful campaign from one market to another is not enough.

Latin America: A stronger two-way conversation habit

In Gupshup’s experience, Latin American markets are particularly active in two-way conversational interactions. Consumers are more proactive about asking questions and seeking information from businesses.

That creates a different starting point for engagement. A business needs to be ready for a conversation, not merely prepared to send a notification.

India: Outbound campaigns remain influential

Beerud contrasted this with India, where engagement has been more focused on outbound marketing campaigns that subsequently lead to conversations.

I asked whether this difference came from customers or marketers. He saw both at work: consumers have become accustomed to receiving notifications, while marketers have become accustomed to running large campaigns.

I would not read this as a fixed description of either market. It is an observation about existing habits, and those habits can change.

Messaging economics shape marketing behavior

Another important factor is cost. Beerud estimated that messaging costs in India were at least five to eight times lower than in Brazil.

That difference changes the return required from a campaign. When communication is more expensive, the business has to be more conscious of how many messages it sends and what those interactions deliver.

He linked India’s historically low messaging costs partly to low SMS pricing, which also influenced pricing across other channels. But he expects rising rates to push Indian marketers toward more personalized engagement and two-way conversations.

My takeaway was straightforward: channel economics and customer behavior need to be understood together. A broadcast-heavy strategy is not only a creative choice. It can also reflect the economics that made that strategy practical.

This connects closely with the broader shift explored in our lessons on conversational commerce and customer engagement: sending a message and helping a customer make progress are different jobs.

4. Voice Notes Are a Product Requirement, Not a Small Detail

The discussion became especially interesting when we reached voice notes.

Beerud described their heavy use in parts of the Middle East, where some customers find speaking easier than typing in Arabic. I mentioned hearing from his team about customers sending multiple voice notes in succession rather than text messages.

For an AI agent, this is not a minor interface preference. If voice notes are how customers naturally communicate, the agent needs to support that behavior, including responding through voice notes.

I also asked whether marketers were sending voice messages proactively. Beerud said he had seen occasional campaigns, but not enough to call it widespread. He noted that initial campaigns often use video, while voice notes can be particularly useful once a two-way conversation begins.

The lesson for me was to separate the format that starts engagement from the format that helps a customer continue it. They do not have to be the same.

5. Omnichannel Means Moving With the Task

Beerud’s final example brought the entire conversation together.

Imagine an AI agent speaking with a customer on a voice call. The customer then needs to provide KYC information or upload a driver’s license. Continuing entirely through voice no longer makes sense.

The agent can send a message, ask the customer to upload the document or photograph, and then continue the conversation.

That is a hybrid experience: voice for explanation, messaging for the next action, and an image or document when the task requires it.

A task-led experience framework

My synthesis of his example is a simple sequence:

  1. Understand the task: What is the customer trying to accomplish?

  2. Use the appropriate format: Voice, text, video, or an image, depending on what is needed.

  3. Switch when necessary: Let the task determine the next interaction, rather than forcing everything through one channel.

  4. Continue the journey: Treat the switch as part of the same customer experience.

This is the practical meaning of multimodal conversational AI. It is not about adding more channels for their own sake. It is about using the right mode at the right moment.

My Biggest Takeaway: Build for Context, Not Just Capability

I started by asking how global Gupshup had become. I came away thinking about how local the next generation of AI applications will need to be.

Beerud’s lessons connect neatly: empower local teams, identify a specific intersection of geography, industry, and use case, understand customer habits and messaging economics, and design conversations that move across formats.

The next big AI opportunity may come from making powerful technology work naturally in a very specific customer’s world. For founders and omnichannel leaders, that is a much more useful starting point than technology alone.

I am Saurabh Agrawal and we come with a new episode on Dilse omni talks every fortnight and cover different aspect of omnichannel with amazing speakers.


Latest Episodes

dilse omni talks nilesh patel

The Lead Management Playbook

Nilesh Patel, CEO of LeadSquared talks about the funnels, customer journeys and the lead management.

dilse omni with swapnika nag

Decoding Personal AI Agents

On this episode of Dilse Omni Talks, Saurabh sat down with Swapnika to talk about what happens when you stop treating WhatsApp as a messaging app and start treating it as the operating system your entire business already runs on.

dilse omni talks with aditya singh

Funding The Next Unicorns

On this episode of Dilse Omni Talks, we covered how to read India's consumer market, what actually happens in a pre-seed pitch room, and why networking is the one compounding skill most founders radically underinvest in.

dilse omni talks with gupshup founder

Future of Marketing is Conversational

Gupshup founder Beerud Sheth on why the click is dying, how conversational AI is rewriting marketing, and why context, not intelligence, is the real moat.

Inside Arjun Vaidya's D2C Playbook

Arjun Vaidya on building, exiting, teaching, and investing, when to go offline, founder branding, and life as an operator turned VC.

View More

Latest Episodes

dilse omni talks nilesh patel

The Lead Management Playbook

Nilesh Patel, CEO of LeadSquared talks about the funnels, customer journeys and the lead management.

dilse omni with swapnika nag

Decoding Personal AI Agents

On this episode of Dilse Omni Talks, Saurabh sat down with Swapnika to talk about what happens when you stop treating WhatsApp as a messaging app and start treating it as the operating system your entire business already runs on.

dilse omni talks with aditya singh

Funding The Next Unicorns

On this episode of Dilse Omni Talks, we covered how to read India's consumer market, what actually happens in a pre-seed pitch room, and why networking is the one compounding skill most founders radically underinvest in.

dilse omni talks with gupshup founder

Future of Marketing is Conversational

Gupshup founder Beerud Sheth on why the click is dying, how conversational AI is rewriting marketing, and why context, not intelligence, is the real moat.

Inside Arjun Vaidya's D2C Playbook

Arjun Vaidya on building, exiting, teaching, and investing, when to go offline, founder branding, and life as an operator turned VC.

View More

Latest Episodes

dilse omni talks nilesh patel

The Lead Management Playbook

Nilesh Patel, CEO of LeadSquared talks about the funnels, customer journeys and the lead management.

dilse omni with swapnika nag

Decoding Personal AI Agents

On this episode of Dilse Omni Talks, Saurabh sat down with Swapnika to talk about what happens when you stop treating WhatsApp as a messaging app and start treating it as the operating system your entire business already runs on.

dilse omni talks with aditya singh

Funding The Next Unicorns

On this episode of Dilse Omni Talks, we covered how to read India's consumer market, what actually happens in a pre-seed pitch room, and why networking is the one compounding skill most founders radically underinvest in.

dilse omni talks with gupshup founder

Future of Marketing is Conversational

Gupshup founder Beerud Sheth on why the click is dying, how conversational AI is rewriting marketing, and why context, not intelligence, is the real moat.

Inside Arjun Vaidya's D2C Playbook

Arjun Vaidya on building, exiting, teaching, and investing, when to go offline, founder branding, and life as an operator turned VC.

View More

Latest Blogs

Dilse Omni Talks — Omnichannel Podcast by Saurabh Agrawal

Talks with Saurabh Agrawal

Dilse Omni Talks — Omnichannel Podcast by Saurabh Agrawal

Talks with Saurabh Agrawal