This Bengaluru AI Startup is Creating the Next Million Jobs for Indian Gig Workers

This post was originally published on this site.

Listen to this story

Last week, when Zomato chief Deepinder Goyal and his wife Grecia Munoz took to the streets of Gurugram, taking upon themselves the task of delivering orders, the idea was to better understand the challenges faced by their delivery agents. Working towards that end is Bengaluru-based AI startup Vahan.ai, which is tackling a broader issue of providing employment opportunities for gig workers aspiring to work with similar platforms. 

Founded in 2016 by Madhav Krishna, Vahan.ai uses AI to tackle hiring challenges for blue-collar workers in India. The platform effectively connects job seekers with appropriate opportunities while assisting employers in identifying the ideal candidates. 

To date, the company has successfully placed over 500,000 workers across more than 480 cities. Among its prominent clients are industry leaders such as Zomato, Swiggy, Flipkart, Zepto, Blinkit, Amazon, Rapido, and Uber. 

Initially, the company started as a platform to upskill people but found it tough to scale, leading to a pivot in 2019. “We realised that it was going to be hard to build a big business in the skilling space, and found that skilling is a vitamin, whereas getting a job is the actual painkiller for people,” said Krishna in an exclusive interview with AIM.

Krishna revealed that they got into Y Combinator the same year. “We were one of 12 Indian companies selected, and raised a Series A in 2021,” he said, adding that, using their platform, companies recruit upwards of 25,000 people every month.

Vahan.ai recently became the third Indian AI startup to raise $10 million in Series B funding from Khosla Ventures, following Sarvam AI and Upliance.ai, which had previously secured investments from the firm. This funding round was also supported by Y Combinator and Gaingels LLC.

All About the AI Recruiter

The company currently offers an AI recruiter that can speak to aspiring candidates over the phone in both Hindi and English. It simplifies onboarding by reducing the barriers of tech literacy. “It handles basic qualifications, checks if the person is interested in the job, if they’re available, etc., and then connects them to a human,” explained Krishna. 

“It asks if they’re interested in a delivery job. It also asks if they have a bike, along with a few other questions to help qualify and match them,” he added.  

Krishna explained that their AI recruiter operates at half the cost of a human caller in India, which is crucial for scalability in a country like India. “The cost of running our tech is 3 cents per connected minute, half the cost of a human caller,” he said.

He further added that their goal is to build an end-to-end AI recruiter that can not only check if people are interested but also answer their questions, match them with different jobs, onboard them, and provide post-joining support. Krishna shared that currently there are about 11 million people registered on their platform, and they’ve placed close to five lakh people.

Creating Millions of Jobs Powered by OpenAI 

Krishna revealed that it is using OpenAI’s GPT-4o in the backend as the LLM, along with speech-to-text and text-to-speech services. The company has established a close relationship with OpenAI. “In fact, we share a strong partnership with them, as they are also backed by Khosla Ventures, just like us.”

He further revealed that they had early access to OpenAI’s latest launch, ‘Realtime APIs’, which was announced at DevDay 2024. “The OpenAI team has established a close relationship with us to help adapt their models for our use case. I’m also in discussions with a few other large LLM providers to establish similar partnerships,” said Krishna.

For instance, the company has developed natural language voice capabilities in Hindi and English with ElevenLabs. The company is also exploring partnerships with Sarvam AI. “We are working closely with them to figure out how we can use their models. Their technology is fairly new at the moment,” said Krishna. 

What’s Next? 

With the latest funding from Khosla Ventures, Krishna explained that the company has two major priorities. “One is to build this AI recruiter and to support eight Indian languages along with a few hundred dialects in the next year. We have very aggressive targets,” he said. 

The second priority is to expand into other industries, such as manufacturing and warehousing. “Right now, we operate largely in the delivery segment,” he explained. He added that while there might be an element of skilling involved, employers in sectors like manufacturing and warehousing face very similar challenges.

Krishna mentioned that Khosla Ventures first invested in the company in 2019. “It’s been almost five years.”  When asked what Vinod Khosla appreciates about the company, he said, “Vinod loves the fact that we are using AI to create a very large-scale impact and that we are highly ambitious as a company.”

Krishna shared that their customers are happy with their service. “Our customers love us and keep requesting more recruits. Swiggy even sent us a congratulatory cake after our recent funding round,” he said.

He explained that in India, it’s relatively easy to build a product and get people to use it, but the real challenge comes when you start charging for it. “What we have built is something that is scaling and showing good economic viability. While we are not yet at breakeven, our contribution margin is positive,” he said.

Currently, the company is focused on India but is open to expanding into global markets in the future. “Beyond India, Southeast Asia is also very interesting. The Middle East is intriguing, and potentially South America as well. So there are other global markets that are definitely of interest, but India itself is a massive opportunity, so we want to stay focused for a few years,” concluded Krishna.