Bengaluru-based Arivihan has raised $10 million in Series A funding from existing backers Accel and Prosus, with a further $200,000 from its existing GSF angel network, taking total capital raised past $15 million. The company, founded in 2022 by Ritesh Singh Chandel, Sonu Kumar and Rushabh Kothari, runs an AI-based learning platform for school students built entirely around interactive video lessons, instant doubt-solving and AI-generated study plans, with no live teachers in the loop at all.
The detail that makes this round worth reading closely isn’t the amount. It’s where the customers already are: Arivihan says roughly 80% of its subscribers come from tier 3 cities and rural India, and the new capital is earmarked for deeper expansion into Madhya Pradesh, Uttar Pradesh, Rajasthan and Bihar, alongside more vernacular-language capability and a wider CBSE offering.
Most EdTech brands built trust the opposite way
The dominant Indian EdTech story of the last decade ran the other direction: build credibility in metro markets with name-brand teachers and polished production, then push downstream into smaller towns once the brand had a track record. That sequencing assumed small-town parents needed more reassurance about a new learning format, not less, because they had fewer alternatives to compare it against and less tolerance for a wasted year of their child’s schooling.
Arivihan’s numbers suggest the opposite dynamic is now in play for an AI-only product with no teacher on screen at all. Where the metro parent’s hesitation is often philosophical - is a machine actually as good as a person - the tier 3 or rural parent’s comparison set is frequently no live tutoring at all, or an overstretched local tuition class with no personalization. Against that baseline, an app that answers a doubt instantly at 9pm doesn’t have to win an argument about pedagogy. It just has to be there when nothing else was.
The real product bet is the AI research line
Burying “AI research” inside a funding announcement usually means marketing language. Here it’s closer to the actual thesis: a doubt-solving and study-plan engine that works in Hindi or another regional language, for a syllabus taught unevenly across different state boards, is a harder technical problem than the English-medium, CBSE-only version most funded EdTech products optimized for first. Arivihan is effectively betting that solving the harder, less-covered version of the problem is also the larger market - and that the trust gap for AI-only tutoring is smaller, not larger, exactly where fewer human alternatives exist to begin with.