It is not easy to run an AI startup in India. Companies and enterprises are reluctant to pay for AI services and products without undergoing numerous trials, checks, and proof of concepts (POCs). This has resulted in several startups starting the ‘Skip India Movement,’ in which sales are not dependent on the Indian market.

However, some startups go to the US for quick monetisation and come back to solve the country’s problems, but AI, arguably, is too early for that. RevRag, a B2B agentic AI startup, decided to undertake this herculean task long ago and is now back in India selling AI to enterprises, even if the payoff takes a bit longer.

“We are not quitting India because we think India is a large market overall, even if it takes time. And we will be at it,” Ashutosh Singh, co-founder and CEO of RevRag, told AIM. 

RevRag is working with the country’s top lenders, insurance companies, and fintech players. With AI-enabled multi-channel orchestration, the startup aims to bridge the gap between enterprises and their customers, ensuring no high-intent prospect is lost due to inefficient follow-ups.

In August 2024, the company raised $600K in its pre-seed funding round, led by Powerhouse Ventures, Kunal Shah (founder of CRED), Viral Bajaria and Premal Shah (co-founders of 6sense), Deepak Anchala (founder and CEO at Stealth AI startup), Vetri Vellore (founder of Rhythms), and over 15 other investors.

Revenue in India Comes in Waves—Not Sprints

Singh acknowledges the challenges of selling in India. The sales cycle is slow, decision-making is layered with bureaucracy, and customers can go cold without warning. “You might get ghosted, and you won’t even know. You’ll keep following up, but the deal might just disappear,” he explained.

Singh estimates that, unlike in the US, where the startup might already achieve an Annual Recurring Revenue (ARR) between $500,000 and $1 million, businesses in India tend to take more time to finalise deals. However, they eventually grow significantly because of high sales volumes. “In India, one must approach work with patience and composure. Revenue grows gradually,” he noted.

The company is currently working towards its first $1 million in Indian revenue, with plans to scale up to $5 million and $10 million by combining India and US operations. “My immediate target is a million dollars from India. After that, the $5–10 million revenue will be a mix of India and the US,” Singh said, estimating a two-year timeline to reach those numbers.

RevRag isn’t putting all its eggs in one basket. It is actively expanding into the US, with POCs already underway with major banks. “We’ll start our US GTM in Q4,” Singh said. “But there will be separate GTM teams for India and the US.”

Still, Singh is clear about his priorities. “Right now, India is our primary market. We flipped back from being a US holding company to focusing on India. We believe that the next 1–2 years are critical as enterprises here mature in their AI adoption.”

One of the biggest myths that Singh wants to bust is that Indian clients don’t pay. “It’s not about inferior tech or lack of money. It’s a game of volume and patience,” he said. “You invest first, like Zomato did, and then you start getting money once the volume kicks in.”

Pricing also emerges as a significant barrier. Even if a POC wins on tech merit, converting it into revenue hinges on aligning price points with business priorities. “The problem is cracking the right pricing at the right time. And you have to solve it with patience,” Singh added.

Despite budget pressures, Singh is clear about one thing: they won’t compromise on tech. “Indians need everything—good cost and good tech. And in that, you have to make it here,” he said. “We will not tweak anything just to cut costs. The product should be good for Indians.”

Despite winning multiple POCs due to its technology, RevRag’s quarterly revenue targets have occasionally fallen short, which Singh attributes to companies taking more time in due diligence. 

While many Indian founders are pivoting to service models due to slow product uptake, Singh believes in a blended approach. “It’s not product vs service. You have to mix both to unlock revenue in India.”

10 People, 20 Clients, and AI Everywhere

Despite being a lean 10-member team, RevRag serves 20 enterprise clients. AI underpins almost every process in the company—from solutioning and prompt engineering to automation and coding. 

“We extensively use AI coding tools. First, we do things manually, then we automate. It’s not just AI; it’s about writing better code and improving processes,” Singh said. He is candid in saying that contact centres will see job losses due to AI. But it’s not all gloom. 

“There are new roles like AI solution engineers, AI testers, and prompt consultants coming up. So while some jobs go, others will be created,” he said.

RevRag builds application-layer AI agents—voice, workflow, embedded—that are trained and orchestrated for specific enterprise use cases. These agents are not built on proprietary foundational models; instead, they leverage open-source and closed-source LLMs, fine-tuned and optimised for tasks such as onboarding, loan servicing, and support in the BFSI sector.

While RevRag is open to deploying homegrown LLMs like Sarvam and Krutrim, Singh notes that latency and production readiness continue to pose challenges. 

“We’ve tried Sarvam, but the latency is high. Once it’s sorted, we would love to deploy it because it understands Indian languages better,” he said. As for Krutrim’s Kruti agent, Singh said they haven’t explored it yet as “it’s too early.”

The company has recently launched two AI agents, Sophie and Emma, both for different use cases. Interestingly, RevRag has intentionally dropped the idea of giving its AI agents Indian-sounding names. 

“If we name it Indian, we won’t be able to sell in the US. If we name it American, we won’t be able to sell in India. So we’ve decided to go nameless and just focus on use cases,” Singh said.

So what sets RevRag apart from IT giants like Infosys, who claim to be building hundreds of AI agents, is the focus and depth of integration. “Nobody in India is doing embedded AI agents the way we are. Our agents can operate in-app and then transfer calls to a human agent, handling the entire workflow,” he added. 

The post ‘We are Not Quitting India,’ RevRag’s $10 Million Agentic AI Roadmap for Two Years appeared first on Analytics India Magazine.