Following the HPCC Systems Summit, I took the opportunity to sit down with Raj Chandrasekaran, CTO and co-Founder of ClearFunnel (www.clearfunnel.com). Raj has a background in technology consulting and has led practices on Technology Strategy, Platforms, and Architecture. Currently on his third start-up, Raj shares information on how ClearFunnel is using the open source HPCC Systems big data platform to solve customer big data challenges as well as advice on approaching big data solutions.
ClearFunnel provides a fresh alternative with an end-to-end and custom Big Data Solutions “as-a-service” model that utilizes HPCC Systems as the preferred big data platform.
For customers, this means:
- Turnkey solutions on low subscription fees (pay for performance).
- Production launch your Big Data use case within 4 weeks!
- With no upfront costs, you do not have to invest in any Big Data talent, technology, and infrastructure.
- Zero risk with try-before-you-pay features and no long-term commitments.
- Shields you from all the complexities of building and maintaining an enterprise class Big Data solution, allowing you to focus on what you do best – growing your business.
Key questions of the podcast include:
0:15- Tell us about your company, your mission and the types of problems you solve for your customers
- Raj discusses the new paradigm of ClearFunnel’s big data Solutions as a Service model and the benefits it provides customers including reducing or eliminating costly consulting and solution creation costs
1:00- What type of data does ClearFunnel typically work with and what data challenges are you encountering?
- From scraping and filing text data to encoded data streaming billions of records per day, Raj discusses some of the challenges ClearFunnel addresses for customers.
2:10- In your journey as a start-up, how did you come to HPCC Systems as your big data platform of choice?
- Having worked with HPCC Systems in his consulting career, Raj shares how he was aware of the platform. Learning that the platform was open source, working on Amazon Web Services, and the same version LexisNexis Risk Solutions uses to drive it’s business made selecting HPCC Systems an easy choice.
4:20- What benefits are you seeing from HPCC Systems?
- As an open source big data platform, HPCC Systems enables the lowest cost development and provides the efficiencies of a mature platform controlled by one simple language, ECL. ClearFunnel uses the HPCC Systems big data platform for many actions including data refinery, data delivery, web services, crunching billions of records, as well as low latency queries
5:25- How easy is it to learn ECL?
- Raj shares an anecdote that if you know more than one coding language, ECL is easy to learn. ECL is a declarative language meaning you just tell it what to do and the language generates the code in the background.
6:25- What advice do you have for companies looking to build a big data solution?
- Raj recommends taking a technology agnostic view rather than starting with a technology stack in mind. Begin with your problem statement and then define your solution while assuming your technology stack can handle the solution.
For more information:
See Raj present at the HPCC Systems Community Day on YouTube Live Stream
View Raj’s presentation deck on SlideShare
You can reach Raj at ClearFunnel in the following ways:
Prefer to read the transcript? Below is the full podcast transcript:
Jessica: Hi, this is Jessica Lorti, and I'm at the HPCC Systems Summit. I'm here with Raj from ClearFunnel.
Raj: Hey, Jessica.
Jessica: Hey. All right, I really liked your presentation. I was wondering if you could tell the folks listening a little bit about your company, what your mission is, and the types of problems that you solve for customers.
Raj: Thanks Jessica for your nice compliments. ClearFunnel is a big data start-up. It's focusing on building a technology solution [00:00:30] . We thought we will come up with a new business model which is big data solutions as a service. What it means is users come to us with problems, we actually create the solutions. We host them on our big data platform and clients pay us only as and when they use them. It's more like a pay-per-use model. Clients are not paying for building a solution, clients are not paying for costly consulting, so they only pay when they use it[00:01:00].
Jessica: All right. Could you share a little bit about the type of data that you are working with, and what are your data challenges?
Raj: One of the main challenges I feel is in the type of data that is coming in. Usually big data, people think it is just text. Or some people think it is large amount of numerical data. But in reality what is happening is the type of data that is coming in is actually our challenge [00:01:30]. One of the clients has come up with encoded data which is streaming out of a normal AAS transmitter which is just 110 characters which by itself has no meaning. But when you look at the total amount of data that is coming in which is like trillions of records per day you start uplaying on our VS relationship and come out with newer insights into what the data is. The type of solution that's good for this is different from scraping five million articles [00:02:00] off the web and trying to do a sentiment analysis. The data is right from encoded data all the way to a text data.
Jessica: Wow, that's a broad spectrum.
Raj: Yes, definitely.
Jessica: All right, so you're a start-up company, and you've definitely taken a journey as a start-up. Can you tell us a little bit about your journey and how you came to HPCC Systems as a solution?
Raj: Sure. I have alternating [00:02:30] backgrounds between technology consulting and product start-ups. This is my third start-up. What we thought is the need today for our start-up is it should be something to do with big data. It has to be something to do with analytics but it should have a unique business model. When we did that our last association with LexisNexis[00:02:52] gave insight into that product called HPCC Systems. When we looked at the HPCC Systems [00:03:00] it was fundamentally fantastic, it was stamped by Lawrence Livermore Laboratory but it was not open source. So we thought we would learn how HPCC works and try to create some competing product. We decided if at all we are building a product we should build it like HPCC. Then two years later we heard that the HPCC is open source. One year later we found out that HPCC is not only open source, it is the same exact edition which is the paid edition, [00:03:30] and it is also available and working in Amazon. With this combination of a top-notch product working in Amazon and it's also big data analytics which is going to be our soft spot made us decide if we're having a start-up it should be HPCC based start-up, and it should lend itself into solving all the big data problems which it is already used to solving. That is how we got associated with HPCC.
Jessica: That's great. You [00:04:00] bring up a really good point. A lot of people don't realize that the open source HPCC Systems is the exact same product for the most part that we use in LexisNexis to drive our almost $2 billion business, and I've had one person say it's like we're giving away the engine to the Porsche so that you can build your own car. It's wonderful to hear you say that. What benefits are you seeing from HPCC Systems?
Raj: As I said our business model is unique that we won't have the lowest [00:04:30] cost where we are not charging people for building the solution but using the solution. If that is the case our cost of developing the solution should be low. The only way I could do it was having a homogeneous stack like how HPCC is built. Today when I'm building solutions with HPCC I get the work of 20 years or what that is gone behind HPCC but all I need to do is learn one single declarative language called ECL. Thinking the fact that [00:05:00] I can do data refinery, I can do data delivery, I can do web services, and I can crunch millions of trillions of data and also do a low latency [00:05:09] query, all because I learned one single simple language called ECL was the primary module of how HPCC's actually helping us. It's hands down that HPCC is going to help you if you want to boot strap your own start-up.
Jessica: All right. How easy was [00:05:30] it for you to learn ECL?
Raj: ECL, when I first went to the class I asked the question to the trainer, "Which language is the one which is closest to ECL? If I knew JAVA is it closest to ECL, or if I know Python or if I know SQL?" The answer which came out was amusing. The answer they gave was, "If you know more than one language ECL is easy to learn." Then I [00:06:00] realized all they meant is, it's not like any other language because they are going for a declarative language. You just say what you want to do and it generates code behind. It's a simple paradigm shift of trying to tell how I will design my solution rather than a normal Java program assisting in designing out solutions programatically.
Jessica: Thank you. How do you, when you work with companies that are [00:06:30] looking to start using big data, what advice do you have for them?
Raj: One thing which I have seen consistently earlier in my consulting world as well as in start-up now, a lot of people start with a technology stack in their mind and then focus on how can they build a solution for it. The knowledge of a technology stack is always limited because you cannot know as much as the creator so my advice would be to come top down. Start with the problem statement, do your solution, make an assumption yes, [00:07:00] the technology stack you already have is going to solve it, and approach the problem as solving a problem with the technology agnostic solution and then focus on how will I design it using my current product. Rather than saying that I know these commands, how do I now solve a problem given to me. That would be my advice.
Jessica: All right, so a little bit of a paradigm shift.
Jessica: All right, fantastic. How can people reach you or how can they learn more about ClearFunnel?
Raj: The best thing is to go to [00:07:30] ClearFunnel.com which is our website, or send a mail to us. My personal ID is Raj@ClearFunnel.com, or they can tweet us @ClearFunnel.
Jessica: All right, fantastic. That's all the questions I have today. Thank you so much for joining me.
Raj: Thank you, Jessica for having me, really appreciate this.
Jessica: All right, thank you.