India's IT scene has been defined by its outsourcing giants. Its future will be defined in large part by companies like dhanaX. I visited the microfinancing startup's Bangalore office this week, just days before its planned soft launch. They've spent about $20,000 to get this far. Can two people, in a shared office with the chairs still wrapped in plastic, really pull this off?

DhanaX is founded by Siva Prasad Cotipalli, a former Oracle India marketing manager, and Prashant Mishra, an engineer who left GE here in India to join dhanaX. They're in the midst of hiring employee #3, a CFO. They quit their jobs early last year to do this, though Mishra has done some contracting on the side to pay his bills. For funding, they've tapped friends and family, and in April took second place in a bank-sponsored business plan competition around microfinancing, landing 1 lakh rupees (about $2,500) in prize money. More important, through that they met an angel investor, who put in more money, gave advice, and let them use an office until they found this one.

DhanaX is a classic example of a startup jumping a hot trend, in this case peer-to-peer microfinancing, aimed at loans too small for banks to make. The concept is people apply for loans online, state their case and background, and individuals make loans them, getting a return, with a dose of feel good for having helped someone. The site launches next week with a very small number of borrowers, whom dhanaX connected with via the Janani Group, an agricultural company in Hyderabad. It's hoping to build more such ties; it would like to tap into ITC's e-choupal initiative, an Internet channel to reach rural farmers. But so does every company that wants into this business. Someone's going to make this peer-to-peer lending idea work, with the promise of eBay-like riches. The attitude around Bangalore, as it is among any entrepreneurs, is why not dhanaX?

There are still big challenges for startups in India, including some cultural trends that work against recruiting employees. With even average employees getting 15% raises a year, and stars getting two and three times that much, Mishra admits it's sometimes hard to think about what he could've done with GE, where his friends are getting promotions and working around the world: "Working for GE was great ... but I wanted to work on something I love. Sometimes I wonder, I think I could be living somewhere else, for example. But I think this is more important."

Some things are tougher for startups in India than in the United States. Simple things like finding a lawyer that's used to working with cash-strapped startups can be tough. But a community's building; in Bangalore, there's a twice a month, Sunday morning coffee gathering of entrepreneurs, where people share ideas and bolster each other's resolve.

But there's also the advantage of being able to keep startup costs low, using a lot of the same techniques Web-based startups in the United States are to make their early costs variable. DhanaX has contracted out a lot of development locally. It's using by-the-drink Web hosting, so it's costs will grow only as it scales. The code is built on free open-source software -- the LAMP stack of Linux, Apache, MySQL, and Perl -- with plans to get enterprise support licenses only as revenue rises.

Indian startups have a lot going for them, from a booming domestic market to a growing entrepreneur ecosystem to a raging sense of optimism that they can succeed. "You get to know what's happening in Silicon Valley, around Europe very easily, so we think 'Why can't we build something like that?' " says Cotipalli. All those factors are necessary to fuel a true Indian startup boom. They're also fragile.