Welcome to the 8th episode of our podcast!

You can download the podcast to your computer or listen to it here on the blog. Click here to subscribe in iTunes.

CP8: Podcast with Wayne Citrin from JNBridge about Java and .Net Interoperability


Martin: Hi. This time we’ll have a very technical entrepreneur on our site. Hi Wayne! Who are you and what do you do?

Wayne: Hi Martin, thanks! My name is Wayne Citrin, I’m the co-founder and CTO of JNBridge.

Martin: Cool. What is JNBridge?

Wayne: JNBridge is a company that creates Java and .Net interoperability tools – it’s the ability to closely integrate code on both the Java and .Net sides so that developers can call .Net code from Java and can call Java code from .Net. We have three products; one is flagship – a general purpose Java/.Net interoperability tool called JNBridge Pro, and we also have two more funereal-focus products, one is a Microsoft Biztalk server adaptor for JMS – Java Message Service , and the other is a more general purpose .Net adaptor for JMS.

Martin: How did you come up with this business idea?

Wayne: The answer is that, at some point, when I was looking for an idea – I had just finished up with another start-up that I was working with. I had a number of conversations with people that were unrelated but in several ways seem to all come together with the related idea.

In the previous things that I had done, I had already done a lot of work with Java and the Java language and the enterprise Java technologies, and a friend of mine here in the University of Colorado had just come back from a faculty trip to Microsoft Research and suggested that I might be interested in looking into this new .Net platform that they demonstrated – this was around the summer of 2001 when .Net was brand new. A little later, another friend who had a start-up of his own asked me whether his enterprise Java beans would work with .Net. This got me into wondering whether or not there was a viable market for interoperability between the two technologies.

Martin: And Wayne, what is your background that prepared you for this kind of opportunity?

Wayne: Before I became an entrepreneur, I was an academic. I did a Computer Science major at Cornel University and I wanted to get a PhD and develop e-expertise in the field and I wanted to be a professor. I like the idea of teaching and doing research. So I went to graduate school at the University of California at Berkeley in Computer Science and I studied programming languages and compilers and in particular, my research was on compiling and executing a language called Prolog, I don’t know if you remember this but it was a very popular language back at the time, very cool, well, it’s called logic programming. It’s still used a little bit but technology has moved on.

In any case, there was not a big start up culture there at the time – which was the early 1980s, what start-up culture there was, was mostly concerned with hardware, this was the beginning of the PC revolution. Some students went to work for companies like Sun, you know, PhD students at Berkeley like Bill Joy, others went to work for Apple. But there wasn’t a big start-up culture – software start-up culture although I did work for my adviser’s start-up for a while. It was in his garage, with other grad students and I worked on the software for a special processor that he was designing that was suited for running Prolog. The customer, I believe, was Westinghouse and I think they wanted to put these Prolog processors on submarines which was kind of interesting.

But at one point, my adviser had to tell us grad students – let them go from the start-up. The University decided that it was a conflict of interest that we could either be his advisees or we could be a start-up enterprise but not both. I really wonder whether this could be an issue today – clearly start-up culture wasn’t the focus at the time but it is now. After the received the PhD, I did a post-doc at IBM’s research laboratory in Zurich, in programming languages and compilations and also network and visualization which were all lot of the focus of the research groups of the lab at the time. Then I got interested in that, I eventually took a faculty position here in the University of Colorado in Boulder and continued the research on programming languages and compilers and the focus was on visual programming languages – graphical ones, it’s a continuation of the work that I was doing at IBM.

After a while, I learned that my passion was really developing software and there was not a lot of opportunity to do that as a faculty member – there was too many other calls on your time: teaching, advising, and administration, so I decided to start a company. It was 1997, the height of the first dotcom boom, and the company was called Lumeo – it was an early website analysis product with an emphasis on visualization and detection of usability issues. I co-founded it with my wife, who has no technical background and has experience in operations, she ran the operational side. We saw a funding and we got a term sheet but at the time, it didn’t make sense to take it. So, what we ended up doing was selling the intellectual property and we used that as the seed money for our next start-up which, you know, I think is the way that things are often done. Closing down the company was a very hard decision to make but it was definitely the right one.

In the meantime, after that I worked in a number of other start-ups in the Boulder area. Boulder, by this time, had become something of a hotbed of start-ups and still is. The strengths here are storage, because IBM had a big plant that manufactured disk drives and there were a number of spin-offs from that but also software connected with IBM also but also the university and some of the other companies that are around here. I think that Boulder is – it is not hard finding people who want to move to Boulder, I mean, between the surroundings and the scenery and the lifestyle, it’s a strong attraction, it really is quite easy, I think, for people to come to Boulder. And right now, like I said, there’s a start-up community, there’s a venture capital community, and so there’re infrastructures there for starting growing companies.

I worked with a bunch of other start-ups after we shut down Lumeo – sold the IP and I learned a lot about what to do and what not to do in starting a company. I worked for a company that was funded with cool technology but had no idea on how to sell it. I worked with a company that was unfunded with no technology and no customers and seeing how they could find something and essentially develop their model. I worked for a very highly funded company that had a business model that was undermined by the Dot-Com Collapse and they flailed around for Plan B and a Plan C and a Plan D, before they finally found something that stuck but they had much reduced the ambitions at that point. At that point, I decided that that company, the highly funded company ended up not doing the things that I was interested in doing so I went off and that point, like I said, it was the summer of 2001, things came together and we came up with the idea for JNBridge.

Martin: And when you started out, how many products have you been working on? Because today, you have a product portfolio of three products but I guess in the beginning you did not start out with all of them.

Wayne: We started out with one product, the flagship product – JNBridge Pro, which was the general purpose Java/.Net interoperability product. But it’s interesting, when we started out with it, I think one of the things that I’ve learned is you come up with the minimum viable feature set for a new product. So currently our product is very powerful but working both directions, you can call from .Net to Java, from Java to .Net. Java to .Net can be running in a same process or in different processes communicating over a network. But at the time, when we started out with was we figured that Java at the time, this is summer of 2001, was a legacy technology so it was more likely you would have existing libraries within Java, and .Net was the next technology where new development was. A scenario we saw as being most common at first would be people developing .Net code and wanting to call existing Java libraries that they already had so our product was simply supported calling from .Net to Java.

Later on, in the second major release, we had a by-directional product – it could call from Java to .Net or actually do both in the same project, also at the time, we decided to make it a socket-based project whereas the Java and the .Net would run in separate processes and communicate through sockets even though I think we really, ultimately wanted to actually have them running the same process because it was so much faster but that was a little bit more complicated and we had to figure out how to best do that.

So we got – with the first version, we actually got a lot of good traction with it because I think we guessed right, you know, people wanted Java to .Net or .Net to Java interoperability and they were happy with the socket-based approach. At the time, I remember we started this project, you know, I went around and I spoke to people who were doing Java and who were doing .Net, and I actually went to Microsoft and the reaction was very different from what we were seeing in developer forums – everybody was asking, you know, “Can I use my Java with my .Net?” and this and that. But when I went to Microsoft and at the time Microsoft had no interest really in interoperability – I would bump to people and they would say: “Oh, why would you want to do that? You would just throw away all your Java and replace with .Net” and I said “Yes, right.” And Microsoft actually I think changed their tune within a year or two because I think most of the market didn’t agree with them. Other people said: “Oh, Microsoft has these migration plans or you could use web services” or things like that but none of these where really good solutions because they weren’t fast enough, they weren’t fine-grained enough. And I think what happened was the market bears this out, those solutions either faded away or they did not cover the whole market and so we were happy with the way that turned out.


Martin: Wayne, let’s talk about the business model of JNBridge. When I look at the product portfolio, is your flagship product still the main revenue driver?

Wayne: That’s interesting. The answer is yes, I believe it does cover the largest portion of the revenue. But the business model is actually in terms of that, had evolved over the years. For one thing when we started it was mainly project-based deals where customers would come and say: “We have a project that we were working on.” and, you know, it’s usually an internal project – they had some in-house library or service that they wanted to integrate. And what’s happened more and more over the years with JNBridge Pro is that OEM deals have become much more important to the business model in the sense that some of them would come and say: “We have a product that requires both Java and .Net and we want to sell this product and we want to embed JNBridge Pro technology in the product either to make it work or to expose features to our users”.

For example, one of our earliest OEM deals actually was with Adobe, they’re called fusion product which is a Java-based product for the last, I believe, it’s probably almost ten years now, has had the ability to access .Net libraries from their Java-based fusion applications and what’s happening is that inside the product, there’s JNBridge technology which they’ve licensed, they’ve embedded in the product.

Other customers want a product where their  end users don’t know about JNBridge Pro because it’s very deeply enough that the features aren’t exposed but if you looked deeply in the product you might see that there’s both Java and .Net and we happen to be the glue that binds them together.

Martin: How is your go to market strategy changed given the change in your customer portfolio so going to more OEMs?

Wayne: I think what happened was that our director of sales actually prioritized this when he found that there was more interest coming in in OEM deals that originally what we did early on, and the customers were happy with this, was that their end users wanted to use features that used our capability or technology, their end users would come to us and buy licenses just to enable those features. But managing that was a little cumbersome and we also had the issues of supporting their end users. Their end users would come to us for support and the problem was that since we didn’t understand the product that it was being embedded in, there were some confusions in making that happen and what we did was we essentially codified this and created arrangements for support where our customers will do the original support and then come to us and plus, we created attractive licensing for OEM deals. I think the market was there from early on but I think we probably started making it friendlier for the users and making it clearer than when this was an option.

Martin: Wayne, when I’m thinking of value propositions, I’m thinking about: “Okay, what are the product or the product portfolio and associated features?”, “What is the gain that you’re delivering to the customer and what is the pain that you are relieving?” And then the next question from me is: “How do you defend this kind of value proposition against competitors? Can you explain – for example, to a non-technical person, what’s so hard about making a service which bridges the Java and the .Net environment?”

Wayne: Well, that’s a great question. At the very simplest, if you’re not really interested in accessing really fine-grained details of the remote service or you’re not interested in high performance, you can probably get away with web services and things like that; there are infrastructures for doing that. But if you want to do things that are very tightly integrated, that access really the entire object-oriented API – application programming interface, of whatever the other platform is, if you want to run the Java and the .Net in the same process. These things actually get pretty complicated. There are a lot of details in managing processes, in managing memory, their various APIs that are involved and are very specialized. And most customers, this is not their expertise, and they don’t want it to be their expertise.

Our main competitor is the idea of building your own. And our customers – some of them actually tried this first and then came to us because they can’t get it to work as seamlessly as they want. Most of the customers, they come to us already have the pain and know that they have the pain, we don’t have to convince them of that. Because if they were able to solve the problem either through fairly simple web service or rolling their own which again would be fairly simple and limited in its capabilities, we wouldn’t see them at that point but once they’ve come to us, I think they’ve already become aware that they have this problem.

Martin: What do you think? How big is the market for the problem that you are trying to solve in bridging this Java and .Net?

Wayne: We’ve looked at surveys by companies like Gartner and Forester. I can’t bring the numbers up in my head or hand but I think that the notions that between 70-80% of enterprises have or 70-80% have .Net, there’s a very large overlap in terms of enterprises that have both, by far the majority of enterprises have both Java and .Net. This could happen for all sorts of reasons: it could happen because they’ve invested in a particular technology or particular proprietary library and one platform but now they want to use another platform; they may have some sort of business logic or financial package written in Java but now they want to have the front-end written in .Net because the UI, the windowing, and the tooling and the graphics are probably better and easier to code in .Net. We’ve had a lot of customers doing a lot in the financial services sector that have that problem.

There are others companies that are highly divided in terms of their IT administration and different parts of the company may be centered on different technologies and they need to integrate. We have customers that face this problem because of mergers and acquisitions; they buy a company and they need to integrate their technologies. It’s a big market, I think.

Martin: For these M&A purposes, are you using, for example, channel partners like consulting companies?

Wayne: Well, we don’t have a formalized partnership program but what happens is that we do deal with a lot of global system integrators. That’s particularly an issue with our Biztalk JMS adaptor, that a lot of projects that involved integrating with Biztalk server are done with consulting organizations. And they will essentially come to us when they have to project or they will actually put together a proposal. They will include our product in their proposal because they know it’s necessary to do the integration. But yes, we do work with a lot of integrators.

Martin: Wayne, can you remember back then when you started out and you wanted to acquire the first customers? How did you attract and especially work with those first customers?

Wayne: Ah, that’s great. What happened was around the time we put our alpha product made available, what we did to start was we essentially took part in developer forums where developers were asking about integration problems and essentially posted announcements that we had this new product that did this particular thing. And, you know, I think there were a lot of people who were kind of sensitive about doing advertising in forums but if you’re providing good information and being helpful and making it clear that you’re with a particular company, I think it shouldn’t be a problem. And what we did was we put those out – that was our initial marketing.

And around January of 2002, we got our first alpha customer. So, it was actually another start-up, I think they were doing something about content management, and it’s been a while so I don’t remember exactly what their technical problem was but it was a developer at the company and I actually applaud his courage in coming to us and taking a chance on us when he had his own brand new piece of software that he was working with.  But what happened was we worked closely with him, we covered most of his technical requirements and his feedback was invaluable on what our 1.0 product was. And of course they got a very good deal from us in return for all the feedback, and in return for being able to talk about this. So that was really what happened.

Like I said, I remember the first customer well, the second couple of customers showed up through the same, sort of, marketing through developer forums, mainly small start-ups and universities. But probably within the first year we were starting to get larger companies and at the time, particularly a financial services. There’s been some interesting questions in terms of our customer base is: a lot in financial services, quite a few in media, broadcast and television-abled companies, there had been quite a few in the oil industry and geographical information systems industry. And in all cases I ask it has to do with the particular requirements of those industries and where the legacy code happened to be written in and I think integrating a lot of that backend code which was probably in Java when you’ve created new front-ends either using ASP dot net or windows forms or things like that.

Martin: Wayne, you started out with this forum marketing. At what point in time did you change your marketing channels and I assume that there was some kind of cap in terms of how many people you could acquired via these forums?

Wayne: Exactly. So relatively quickly, what we did was we concentrated on web-based marketing search engine optimization and placement Google Adwords and those continued to be our main marketing. But there are other things we – that happened fairly quickly, early on in that first year.

Another thing that happened which actually I think worked out quite well was connected with the forum-based marketing, a couple of people at Microsoft were working on interoperability issues contacted us and one of them, a guy named Simon Guess, who is no longer in Microsoft but at the time was, started creating internal interoperability, sort of, seminars. He conducted a couple of seminars at Microsoft and invited people who were working on this issue to present, that actually helped us create more awareness in the Microsoft organization. And similar things happened at Sun although it was a little bit slower. Simon also started writing a book about interoperability and published in Microsoft press and we participated in that. And again, within a year or two, Sun wrote a book too and I think some of that added a large contribution to spreading the word.

Other things that we’ve done over the years: we published articles and provided briefings and interviews to journalists and analysts and created sponsored content and these definitely have contributed but I think the real way people found out about us is they type in Java/.Net integration or Java/.Net procedure calls or Java/.Net interoperability into Google, and our name comes up high in that list both on our own website and in articles that have been written about the topic.

We also used to exhibit trade shows and conferences but we don’t do that anymore because, you know, I think there’s not enough bang for the buck in doing that – we did get some good publicity but I think at this point we have the critical mass that we can do without that.

Martin: Yes, great.


Martin: Wayne, you have started several companies and over the, let’s say, 10-15 years, what are your major learnings that you can share with other people interested in starting a company, especially the do’s and don’ts or failures that you’ve made?

Wayne: The first thing that I would want to say is that: if you can, self-fund your company, if it all possible – raising money can become all-consuming. It’s much better to spend the time building and selling your product. And also taking money, in some ways, make you less agile, there’s certain otherwise viable opportunities that you might want to enter or engage in and they might be too small to please the investors, in some ways if you’re not answerable to investors that actually can help in starting your company.

I mean there’s this one story I remember reading that was written by Paul Gram where he talked about this company that eventually became Yahoo stores, he said that when he heard that his competitor had taken a big investment, he knew that he had won because at that point his competitor was hamstrung and really couldn’t profitably engage in the kind of opportunities that Paul Gram’s company could because the competitor had to grow faster and he tried to engage in bigger opportunities.

So what happens is that if you do sell fund, it will affect the kind of company that you can become because of course, your business model can’t be too infrastructure-intensive. Clearly, you know, if you wanted to build data centers or build out consumer hardware, things like that, you may need much more of an investment but for enterprise software that like what we’re engaging in, that was not the case.

Use lots of cloud services, it’s another thing that self-funding will force you to do. Everything from your email to your VPNs.

Don’t do it yourself if you can help it. Also, it forces you to stay lean for a long time; you don’t ramp up your staffing ahead of revenue unlike a company that takes up VC investment.

Virtual offices. We run on a virtual office, I think that’s actually saved us a lot on in terms of worrying about real estate and being a tenant, our company is kind of scattered in various places, we let people arrange their own work environment and that worked just fine for us. But all these essentially add up to limiting your fixed cost, so I think that, but again self-funding is probably the most important decision that we made in starting up.

Another thing that I think I’ve learned is that starting a company is more than a full-time job and you need to get used to that. You think this is obvious but I met a lot of people who think they can start a company on the side while they’re doing something else to pay the bills. In my experience, this really can’t be done at least not well – you get distracted, you work on, you know, what’s bringing you the money, and you know, things just don’t happen with a company you really don’t start-up so.

When we started out, we thought that maybe we would with JNBridge, we thought that we would do what it takes. If there was consulting, we would do that, but it turned out that once we got the opportunity to do some consulting, not necessarily related to the product we were developing, we decided to turn it down because it was too much of a distraction – I think that was the right decision to make.

If you need to hedge your bets, I think you should try to arrange this with your life partner or your spouse can provide financial support well while the company gets off the ground but I have to admit that we did not follow that advice ourselves – we put all our chickens in one basket and my wife is the co-founder and betted up in non-development aspects and I think that worked because we believed in what we were doing and we really felt that there was a good – a much better than even better chance to succeed.

I also think that you really ought to know the field you’re working in, inside out and love working on it. You don’t have to give a lot of time to get your product out the door, and you can’t spend it learning too much new stuff but when you are learning, you really have to enjoy learning. And when we started developing our first product in July of 2001 and I guess the alpha was available in December of 2001, January of 2002, data by March, 1.0 by June. You really have to be in a hurry which means you have to know what you’re doing and also means you have to focus and I guess that goes back to the idea of not being able to just do this on the side or – it has to be your focus and what you really want to do every morning.

We’ve been doing this for almost 15 years, at this point, of course, we can enjoy life more than we had but even back then I think we were having a lot of fun but were really concentrated a huge amount on getting the company started.

Martin: Great. Wayne, thank you so much for sharing your knowledge.

Wayne: Well, thank you Martin.

Martin: Welcome. If you’re thinking about optimizing your code interoperability, check out JNBridge.

Wayne: Thanks.


Thanks so much for joining our 8th podcast episode!

Have some feedback you’d like to share? Leave a note in the comment section below! If you enjoyed this episode, please share it using the social media buttons you see at the bottom of the post.

Also, please leave an honest review for The Cleverism Podcast on iTunes or on SoundCloud. Ratings and reviews are extremely helpful and greatly appreciated! They do matter in the rankings of the show, and we read each and every one of them.

Special thanks to Wayne for joining me this week. Until next time!

Comments are closed.