Voicing AI Loves India, But Knows Startups Still Struggle Here

When Abhi Kumar founded Voicing AI in 2024, his experience with markets’ response to emerging technologies came into play, particularly how to deal with the initial resistance.

He had managed Microsoft’s emerging market strategy globally, co-heading its AI investment fund M12, and having deployed over $1.2 billion across 120 companies between 2019 and 2023. 

Back in 2014, when Satya Nadella took over Microsoft, Kumar remembers questioning the company’s leadership: how many of you have sent a WhatsApp or WeChat message in the last 24 hours? Half the hands went up, mostly among employees from India or China. 

“Americans had no idea about WhatsApp or WeChat. They used text messaging,” he said. That insight eventually shaped Voicing AI. It was there he began forming a hypothesis: AI could play the same role in this decade that browsers did in the 1990s — an interface unlocking entirely new business models.

“When the first browsers came out, they didn’t hold value themselves, but connected people to the internet and enabled PayPal, Amazon, Netflix. AI, to me, is the next browser,” said Kumar in an exclusive interaction with AIM.

The Overcrowded Agentic AI Space

Voicing AI builds synthetic agents that can take calls, chat, respond to emails, think through customer issues, and complete tasks — essentially replacing the lowest tiers of call centre work (L0 and L1), while escalating complex cases to humans. 

By March 2024, Kumar said, they were passing the Turing test in more than 90% of cases. Still, technology is only half the story, the bigger challenge is figuring out where AI fits into the messy reality of enterprise operations. 

Kumar leans on his venture experience to tackle this: “It’s not the best tech that wins, it’s the tech that’s implemented to deliver business outcomes.”

But even so, the market is getting crowded. More and more agentic AI startups are coming into the space. Players like Gupshup, RevRag.ai, Yellow.ai, Sarvam, and other AI startups have been racing to sign enterprise customers.

This is because most of the agentic AI traction is in the B2B space and not for consumer apps. Startups are racing to stake a claim, but most are building on the same underlying tech stacks, fine-tuning existing LLMs rather than creating defensible IP.  

Read: India has 109 Agentic AI Startups Building in a Vacuum

One of Voicing AI’s competitors, Gupshup has been aggressively expanding its AI offerings, moving beyond messaging APIs to voice bots and customer engagement. Founder-CEO Beerud Sheth had told AIM that he had also been preparing for the agentic AI wave. 

Sheth believes that real value lies in the application layer. That’s where Gupshup’s existing relationships with enterprises become its biggest asset. The same goes for Voicing AI.

In December last year, LTIMindtree invested $6 million in Voicing AI for its proprietary technology that brings human-like voice capability across more than 20 languages, while using all of the open source tools in the market and fine-tuning those for specific use cases.

This demonstrates the importance of securing enterprise adoption over having a sophisticated model. It’s why companies from messaging veterans to new AI-native entrants are crowding the same field, often targeting the same set of early enterprise adopters. 

In such an environment, speed of implementation and integration into messy, real-world workflows becomes the real moat.

PoC Purgatory and Meagre Margins

Still, as the market crowds, the pressure to win large contracts — especially in higher-margin geographies — will only grow. Indian enterprise deals might help with credibility, but they won’t drive the kind of revenue needed to outpace global rivals.

Kumar believes the economics favour AI agents. In the West, labour costs are high, attrition in call centres is a constant problem, and multilingual coverage is expensive. AI can work 24/7, deliver consistent quality, and switch between languages instantly. 

In theory, India should be the perfect market: English is widely spoken, the outsourcing industry is massive, and companies are under pressure to cut costs. In practice, it’s the same story that has played out for SaaS, AI, and nearly every other enterprise tech sector — Indian customers love pilots, but rarely pay.

Kumar finds Indian enterprises to be surprisingly “very very AI-first” in mindset, even if their budgets don’t always match their enthusiasm. “I truly believe AI is the era where India has an intrinsic advantage globally,” he said, pointing to the hybrid human-in-the-loop model that many companies will operate under for years to come.

The economics, however, remain challenging. In the US or Europe, replacing or augmenting a $35,000-a-year customer service agent with an AI agent offers plenty of margin to work with. 

In India, where an equivalent role might pay ₹3 lakh, the cost savings are less compelling, and some companies still opt to “just hire five engineers” rather than pay for automation, said Kumar. 

This leads many Indian AI startups to look abroad for their highest-value customers — a trend that SaaS companies experienced a decade ago.

The problem isn’t new. For over a decade, founders have grumbled about the “PoC purgatory” that plagues Indian enterprise sales. The pattern is familiar: a startup spends months building a pilot for a large company, often heavily customised, but gets no contract, no revenue, just a “we’ll get back to you.”

Read: Free PoCs are Killing Indian AI Startups

Kumar advises AI founders to think globally from day one, building for the markets where the unit economics make the most sense.

He hasn’t written off India though. Some of Voicing AI’s largest customers are here, and he sees the market as a testing ground for operational complexity — multiple languages, high call volumes, and customers who expect fast, personalised service. If a system can handle that, it can handle almost anything.

The post Voicing AI Loves India, But Knows Startups Still Struggle Here appeared first on Analytics India Magazine.