Sunday, March 28, 2010

On innovation by Indian IT firms - comment to TechCrunch

There was a TechCrunch post about workforce education where the author claimed that India does it better. I posted a long comment which is posted below. The main point is that India has a lot to learn from America on playing in the right part of the value chain. Training people to be better cogs in a wheel will get India nowhere (fast or slow).

You need to read the original post first.

Here's my comment:

=========

IMO there’s a huge fundamental flaw in your analysis. Firstly, I do agree that companies train less now and expect people to produce immediately. That’s simply reality globally – and the reason Indian companies train a lot up front is often because most new hires are unable to produce much (any?) without such training.

But that’s not the main issue I have with your analysis. You write:

“Additionally, the Indian R&D industry has been moving into the higher realms of innovation. In the aerospace industry, Indian companies are designing the interiors of luxury jets, in-flight entertainment systems, collision-control / navigation-control systems, fuel-inverting controls, and other key components of jetliners for American and European corporations.”

Well the R&D industry may be moving into “higher realms of innovation” but *whose* problems are they solving? Did they think of those problems and say they’re going to find a solution? As you yourself said they’re doing it “for American and European corporations.”

Therein lies the rub. All they’re offering is low cost paid labor. They are not scratching an itch they have. They are not innovating – they are simply delivering value to someone else in raw form so that they can refine it and make the real money.

That is what we in old colonial countries do so well – take our raw material and export it so that someone else can refine it and send it back as value added products for us (and the rest of the world to consume). I don’t need to tell you where the real value lies in that chain.

All you’ve described is that Indian now not only answer phones and transcribe documents (essentially low skill labor) but that they also solve hard analytical problems and do lab research for “American and European corporations”.

That is not innovation – that is just slave labor. Yes that’s right – to me that is simply selling bodies at whatever margin the Indian bosses can make something at.

Of the areas you mentioned, how many Indian brands exist in the world? How many products have Indians thought of and take it all the way to market in the west? Until India starts doing that, there’s nothing for America to learn from India. Right now, there’s TONS for India to learn from America.

India needs to stop selling its people and start finding problems and opportunities and start addressing them and competing globally with innovative products and services. That is the beauty of the US – that there are always new entrepreneurs who pop up with brilliant new ideas who go on to challenge & change the world. With every 100 of those there is one winner. If you don’t have hundreds popping up every year, the winners won’t be there. Of course there are a few exceptions in India but those numbers won’t be enough to really become a global player in the innovation driven IT world. (I know nothing about other industries so I am not going to comment.)

The real problem is that the Indian bosses are happy with their 30% margin. They’re happy because they can make billions more by hiring a few million more Indians to scale up. Ah but have you done a study of per-employee revenue generated by Indian BPO & services companies vs. major IT companies? My understanding is that its in the range of $30k/year. IBM does about $250k/year ($100B by 400k employees), Microsoft about $500k/year ($60b by about 120k employees) and Google about $1m/year ($20B by 20k employees). Do you think that doing R&D for “American & European corporations” will ever get Indian companies to the Google levels? Not a chance.

I’m a Sri Lankan who lived in the US for 16 years and returned home 8 years ago. In WSO2, the company I started with Paul Fremantle, we don’t treat our Sri Lankan team as “cheap labor”. They are part and parcel of the innovation engine we have created. In fact, they *are* the innovation engine we have created. We are now the only truly 100% open source alternative enterprise middleware platform to IBM, Oracle & Tibco. And a lot of the innovation came from young kids in Sri Lanka. We’re about to launch our entire middleware platform as a “Platform as a Service” offering – totally out innovating and going ahead of everyone.

Pretty much the only difference between us and a bunch of Indians in some Indian company is the mindset we bring to the table. We don’t view anything as “someone else will define the problem and we have to solve it”. Rather we look at it as “how do we do MUCH better than IBM/Oracle/Tibco and all the little guys out there”. We encourage and support people in thinking disruptively. We teach each other. We listen to the world. We reward risk taking. We give the freedom to work and think in a way that enables innovation.

With all due respect Vivek, you and other business school types are doing Indians a huge disservice when you tell them “bat on” with what’s going on. If India wakes up and starts challenging the world with innovation that originates from India then there will be something to talk about. Telling the world that India has cheap labor from low skills to PhDs is something I’d personally be ashamed to talk about rather than be proud of.

Sanjiva Weerawarana, Ph.D.
Founder, Chairman & CEO; WSO2, Inc.
Founder, Director & Chief Scientist; Lanka Software Foundation
Member; Apache Software Foundation
Director; Sahana Software Foundation
Visiting Lecturer; University of Moratuwa

Wednesday, February 17, 2010

WSO2 launches Cloud Identity service

Today we announced our second cloud service: Cloud Identity. See: http://wso2.com/cloud/services/identity and you can use it right now at http://identity.cloud.wso2.com/.

(We have a principle of not announcing vaporware!)

This is basically our WSO2 Identity Server product converted into a multi-tenant identity management system and hosted on Amazon EC2 for scalable and reliable deployment. Basically, within about 5 minutes you can register your domain, add your users and then have your own OpenID, Cardspace, SAML 2.0, WS-Trust STS for authentication and XACML and (very soon) OAuth for authorization. Translated to English, that means you can get  a single place to manage your users and give them access to Drupal, Liferay, Google Apps and a whole lot more. We will soon be adding LDAP to this list as well, which means you can even tie Windows, Unix login to it as well as other services like SVN.

Nothing like giving it a try to see how it works!

[UPDATE] Here are some additional references for you to get started with:

Happy identitying!

Sunday, February 7, 2010

The myth of rogue states

The February 8th issue of the Newsweek (International) magazine has an absolutely great article titled “End of the Rogue”. The article is about how the concept of a “rogue state” (apparently created the cold war days) is no longer valid and how the US needs to get past it.

Not surprisingly many comments on the online edition don’t agree that the US approach needs to change. Living in Sri Lanka, however, and having observed the wrath of the US (and UK and EU) for the way the anti-LTTE war was conducted and ended, I can see what must be going on in “rogue” countries!

The most interesting quote I found in the article is this:

We don’t have the right to think other people should think like us.

If we all could live by that the entire world would be a better place!

Wednesday, February 3, 2010

WSO2 platform overview

We recently posted a slide deck that gives an updated overview of the WSO2 platform. This covers both our downloadable products as well as our cloud offerings. Enjoy!

Saturday, January 2, 2010

Delivering a complete middleware platform under the Apache license

Let me start by wishing everyone a wonderful 2010!

Right from the get-go, WSO2 was designed to be a company that built a complete middleware platform. We set out to target the big guys who have a complete story, except with two key fundamental differences: our technical approach and our business model.

Our technical approach is of course based on Web services and SOA. For the first time in the history of computing, Web services have offered a lingua franca for how systems interact with each other. There were of course many previous attempts, but one camp or the other of the technology industry didn't agree and so there was no "English" of the computer world. Web services has changed that with every major and minor vendor supporting interoperability via Web services (XML, HTTP, SOAP and the rest of WS-*).

SOA, despite the much ballyhooed story of its demise at the beginning of 2009, is not only alive and well, but is in fact kicking butt. SOA is fundamentally an approach for how to build large scale composite systems. As an approach, it mimics the real world's service-oriented economy. As such, SOA is a fundamental concept, not some vendor-driven theory. That said, SOA, like any other technology, has had to live through the Gartner Hype Curve. If at all instead of 2009 being the year SOA died, it became the year it came out of the trough and started climbing up towards the plateau of productivity.

Of course the fall into the trough was not without reason for SOA and Web services. Much of it was driven by middleware vendors not delivering anything new, anything valuable in the form of SOA middleware. Many of them simply took their existing middleware and rebranded it the shiny new SOA gimmick. Well that of course doesn't work and the cracks in the story will force you down to the trough .. and it did.

WSO2 is unique in having started from nothing and set off on a path to build a complete middleware platform with Web services and SOA in its heart. The result is simply orders of magnitude less complexity, much better performance and overall greater productivity and lower TCO. These are not random claims from me - these have all come from our users and customers.

We now call it Lean Enterprise Middleware. Try it and see - you'll be shocked at how lean it us, how productive it is and how much money you can save by replacing your legacy or pretend open source middleware stack with ours.

Now let's talk about the business model. Right from the beginning, we made a strong commitment to releasing all of our software under the Apache license and to not attempt any bait-n-switch type acts. Believe me, that took a lot of hard work to keep going .. investors for example have a major issue with the Apache license. Why? Well because you can take any of our software and do whatever you want with it and never ever pay us. We have no legal recourse to making you pay (as dual license business models do) nor any way to force you to pay for the good stuff (as many "commercial open source" companies do). Instead, we rely on delivering real, measurable value to our customers without forcing them to pay us. Our customers love us because they pay for the value we deliver to them, not because we are using the law to force them to pay for the software they use.
When I say you can do whatever, I mean whatever - recently one of our competitors sold a support contract for one of our own pieces of software! Yes, that is possible. In this case the people who will pay the eventual price is the customer who did the stupid thing of buying support from someone who has nothing to do with the software! Remember Oracle's Unbreakable Linux? Well that didn't break Redhat and neither will this act - it just shows how low some people will go to make a buck.
So today you can download an entire enterprise middleware platform from us without registering, without paying, without any risk of bait-n-switch for absolutely no cost. How can we afford to do that and become a successful business? We have many many customers who happily pay us to provide maintenance, provide help and in general to be their technology partner. So having thousands and thousands of free non-paying users is not a problem for us - that's free marketing and helps us save the world from the ugliness that is IBM, Oracle, etc. middleware.


WSO2 is delivering on the promise to build lean enterprise middleware and deliver 100% of it as open source under the Apache license. Oh yes, we also offer it as various cloud offerings - virtual machines, or online services.

We are the ONLY vendor offering a complete enterprise middleware platform 100% open source under the Apache license.

That was all wrapped up in 2009, a tremendous year for us. In an environment of economic uncertainty, not only did we meet our targets but we beat them. We have been doubling revenue each year and this year was no different. We are on a roll :-).

Looking towards 2010, we have more work to do to make our enterprise middleware platform simply untouchable by anyone else. We're already far ahead of our competitors with our WSO2 Carbon powered platform, but we have several things planned to further leave our competitors in the dust. As I wrote in an earlier blog, we practice open development - so if you want to be part of it come on over and join us on architecture@wso2.org!