- Wealth PMS
A company run by two close friends, iCreate Software has just secured Rs. 15 crores in funding from IDG Ventures. Based in Bangalore, iCreate makes business intelligence software for banks. I’ve had long and fruitful discussions with Anup and Vivek, the co-founders, about the future of BI and Banking – it’s such an enormously huge sector, and the technology adoption at the smaller banks is staggeringly low. A market waiting to be tapped, and the competition is looking at the more lucrative US/European banks which leaves the field more open in Asia/Africa/Latin America.
Domain-specific BI is the way to go, according to IDG MD Meenakshisundaram, and I agree. I knew folks at Manthan Systems, a BI Provider for the retail space, and even there the scope is enormous. Subex has niche BI solutions – from Fraud detection to Route Optimization – for telecom providers, and a public company.
In Banking, BI has been of very huge interest to me; in India you see very little data on bank portfolios, loans, credit ratings etc. (at least, the kind they reveal at quarterly meets in the US, like this Freddie Mac quarterly result supplement)
Plus, there’s a new field out there – market impact – that banks should be looking at. For instance, if the CDS spread on a certain company expands suddenly, and a bank has a big exposure to that company, it should be worried about the quality of that asset. Stock prices affect market value of collateral – in one case I know, a bank could only update it’s collateral values for loan-against-shares once a month because the manual process took so much time. While it’s all okay nowadays to remove mark-to-market and therefore hide the real asset quality from shareholders, banks must internally need a better and more real-world view.
Look also at the stupid kind of regulations we have in India – KYC rules for instance – which make us believe we can detect fraud or money laundering. True frauds will easily circumvent the system, while honest players get inconvenienced. Frauds operate in easily identifiable patterns, so does money laundering. You can detect such using better BI, by mapping or crunching data. For instance, someone who wants to launder cash may do a series of cash transactions, just below the 50,000 limit that institutes PAN card verification, or deal with a few parties over and over again. A stolen ATM card may show sudden splurges, that could trigger flags and phone confirmations. Personal loan or credit card defaulters could have triggers from stock markets, regional/sectoral events or even lack of bank account credits.
Having data mining/analysis can solve these problems, and you won’t get anywhere if you build a “generic” data mining solution. It’s like handing me an Intel CPU – I might know how to get a motherboard, memory, disk and cabinet and put it all together but it would save me a lot of time and money if someone were to sell me a full PC instead.
Getting BI currently is a hugely expensive and time consuming operation – typically banks go with a BI horizontal like Cognos, get a system integrator like EDS, IBM or Infy to copy data over from their core banking platforms to an OLAP back-end, and then spend time specifying the reporting so that per-head-billed consultants can build them in Cognos. First, this is out of reach of most smaller banks, and second, the management face time plus the cycle time is just too much.
iCreate’s been approaching this in a very interesting way; they have adapters that can get data from core-banking systems, and they have pre-packaged reports in their BI system. Both of these reduce the cost and cycle time enormously, and lesser time needed from management side to specify reports everyone will need. (Think of canned Basel X, or country specific regulator reports) Reduces time to sell, implement and get a customer happy – and defects from the model of “we need more people to make more money” approach so common in outsourcing.
Enough said. I know this is hugely different from what I’ve said on the blog but still: