From Potholes to Property Titles: India's AI Start-Ups Tackle Everyday Crises
A new wave of Indian AI companies is applying the technology to distinctly local problems, from disputed land records and damaged roads to overburdened public hospitals, even as chip shortages and weak customer spending threaten their growth.
By Amit Ricci ·

In 2020, Sanjay Mandava travelled from Chicago to Chennai in southern India to settle a family land dispute. What followed was months of shuttling between government offices to gather a dozen records proving the property belonged to his family. A sale deed from when his parents bought the land was not enough, and authorities demanded a string of additional documents.
The frustrating experience pushed Mandava to build Landeed, an AI-led digital platform that lets users pull together fragmented land records from multiple sources, verify titles and run legal checks in a few clicks across 24 Indian states. The problem, he says, is uniquely Indian. Unlike in the US or UK, where the government verifies ownership when a property is purchased, India follows presumptive titling, meaning the buyer must establish ownership. A sale deed only records the transaction, and documents such as building certificates, property tax receipts and land survey maps may all be needed to prove title.
The stakes are high: property disputes make up the bulk of civil litigation in a country where roughly 80% of household wealth sits in real estate. Landeed has mapped large parts of India using geo-spatial technology and data from state land revenue departments, registration offices, survey records and municipal databases, linking registered properties to official land identification and survey numbers. An AI layer reads titles, extracts information and flags inconsistencies within minutes, work that takes lawyers and brokers several days. With 1.8 billion documents, the largest title data repository in the country, the app now serves millions of individual and enterprise customers during real estate transactions.
Landeed belongs to a growing crop of Indian start-ups using AI to attack complex problems in legal-tech, healthcare, infrastructure, education and agriculture. While the US and China dominate the building of frontier language models like ChatGPT, India's AI revolution is unfolding downstream, in data centre infrastructure and consumer applications, often trained on highly fragmented, multilingual datasets in extremely cost sensitive industries.
In Bengaluru, YouTube teacher Gaurav Sen built an AI app that detects potholes using a dash-cam, satellite-based navigation and an AI vision model, then identifies the civic officials responsible for repairs and sends them the data. The app went viral recently, reflecting how widespread potholes are across India's urban sprawl, and at least half a dozen similar services now map and report road damage to authorities nationwide. "A municipal corporation can take the app and use it on public buses and taxis, to access real-time, updated data on road conditions," Sen told the BBC.
In Mumbai, Fractal Analytics, among the few listed Indian AI companies, has rolled out its AI tool Vaidya in hospitals run by the local civic authority. The app gives citizens simplified, AI-generated summaries of lab reports in multiple languages, points them to nearby healthcare facilities and offers doctors AI-guided pre-consultation support, such as patients' medical history and medication details. It also feeds hospital administrators a public health dashboard built on aggregated, anonymous lab report insights, giving them early visibility into emerging infections and disease trends.
Investor interest has followed. Funding for AI start-ups jumped 265% year-on-year, according to start-up information platform Inc42, with more than $1bn raised in the first nine months of this year despite a relatively muted fundraising environment for the broader ecosystem.
Yet turning pilots into viable businesses remains an uphill climb. "The cost of building in India is sometimes much higher than what Indian customers are willing to pay," Fractal co-founder Srikanth Velamakanni told the BBC, noting that computing power costs the same whether a model is built in India or the US, while American customers can pay far more. Founders also face a scarcity of high quality chips at the right scale and price. "I believe I am building one of the most impactful things for this country," Mandava says. "But I am starved for GPUs. I have a hellish time trying to acquire them." The government has promised start-ups 38,000 GPUs through its AI mission, but demand far outstrips supply, and cumbersome regulation, bureaucracy and a shortage of experienced talent add to the strain.
Experts warn that unless these fundamentals are addressed, Indian AI firms risk building shallow solutions that the next generation of founders will quickly make redundant. Sen is blunt about his own product: "Someone like a Google will probably laugh if I tell them of my pilot. To make it sustainable and hard to replicate, it will have to be much more sophisticated." Mandava, asked what he wants from policymakers, is equally direct: "We can still do it. We don't want protection, but propel us, give us the weapons to compete against the best." Quoting Jay-Z, he adds, "I can't help the poor if I'm one of them."






