This post was originally published by Tamar Weinberg in March 2016 in HostAdvice.com. The article name is Cartika is Your Application Service Provider on an IaaS environment Tamar: I love hearing stories of people who realized they preferred working for themselves, and even going on soul-searching trips to find where to go from here. I also love companies in the hosting space that are especially innovative, who aren't selling the standard shared+reseller+VPS+dedicated hosting packages but who find ways to clearly be differentiating businesses. That's why it was awesome to sit down with Andrew Rouchotas of Cartika who told us all about his journey. Tamar: Hi Andrew, can you tell us a bit about your role at Cartika? How did you get started and how did the company from? Andrew: I'm the original founder and owner of the company. I do a bit of everything and I focus on leading the support and sales and marketing teams and getting engaged with the finance guys to make sure we're making the right financial decisions and positioning. I take a high-level role but with a company of our size (30 employees), I wear a lot of hats which is the nature of the beast. Back in the day, about 15-16 years ago, I was working at IBM and identified a mid-market SMB opportunity to provide application service provider (now called SaaS) to go to market and sell hosted applications. Most of the time, these were ecommerce solutions on open source platforms. Back then, osCommerce was the standard, and we sold ecommerce solutions to companies where we did design, implementation, consulting, and training. The main issue with running that business was that we couldn't find a good managed hosting provider. So eventually, we started co-locating our own equipment and managing our own environment and hiring our own sysadmins. Our business ended up transitioning because there was demand for our managed hosting services. We stopped focusing on applications directly and started focusing on managed environments and on reseller channels. Tamar: What services do you provide at Cartika? Andrew: Today, we've transformed again. The world has changed from the more traditional hosting into two levels: infrastructure and associated services. Cartika really focuses on those two aspects, on providing pure infrastructure and infrastructure as a service. A lot of people still use the word "cloud." We provide various levels of management and services on top of that infrastructure for customers. Tamar: Can you explain what you mean by IaaS cloud servers? Andrew: IaaS is the "cloud," essentially. It's providing infrastructure and resources in a service format. Our consumers come and buy the infrastructure to meet their business requirements. SaaS is an extension of this. A SaaS provider would typically work with an IaaS provider for the infrastructure and they'd layer their application on top of that infrastructure. "Cloud" is the buzzword that drove the industry toward the utility-based model. With IaaS, we're providing the resources, abstraction layer, networks, virtualization layer, and the IP space, and the customer would provide the internal data/services to the cloud. Tamar: Can you tell us a bit about the solutions you provide for healthcare and the financial space? Andrew: We're really focused on those two aspects: the infrastructure and the support for pure infrastructure services. We'd also provide services like complementing the IT staff or offering services such as managed firewalls or managed backups. With healthcare, you have to worry about additional standards such as HIPAA, and in finance, there's FISMA. Those compliance requirements come with certain expectations. We bundle those requirements for that industry and our infrastructure supports this within the compliance needed and the certifications necessary. Tamar: How does Cartika embrace developers through its offerings? Andrew: Typically, what we find is that developers built a platform and need various levels of solutions that serve their customers. We try to work with developers or development firms on what they need. For example, some small customers may need managed cPanel environments, and others may have a SaaS product and need a pool of resources and an IaaS farm. We created an environment and tools and processes to roll these out more quickly and efficiently for developers to onboard more customers more efficiently. Tamar: Tell us more about Hsphere. Andrew: Hsphere is a Parallels product and was kind of our original platform. It's an engine that we use in our managed environment. It supports clustering natively as well as Windows and Linux installations. We can deploy server farms and allocate resources over a single account. It's a legacy offering for us now. Tamar: What is your accounting offering? How did that evolve, since it seems so different from other offerings provided by Cartika? Andrew: This is a partner product offering. We started with an application-centric focus and are more in a managed services focus. It's really difficult to support all applications, so the way we do this is through partnerships. If an organization requires a more managed type of product like managed WordPress or managed ERP or cloud-based IaaS solution, we'd send them to one of our partners. If it's a financial company and they'd need something built on an Accounting application that we don't provide, the partner would fire that up on their platform and deal with the customer directly but we'd be providing the infrastructure and all sorts of support, management, and FISMA compliance for example as needed. All of this goes through our partner channel. Tamar: I see you have an API. What made you take that direction, and what applications have you seen evolve from that offering? Andrew: The API is absolutely critical especially as we move forward in IaaS and in cloud computing. It really allows organizations to manage things centrally. What we're finding is that organizations and partners will set up resources on various providers and still want to use our services. Therefore, we allow them to use the systems internally within their organization to manage resources and other account functions on our side. We can manage things like contracts/invoicing, but they can deploy and reclaim VMs and associated services/resources to support their business processes. It's a really important resource in the infrastructure space. We have a pretty good sized SaaS provider who has their own portal for managing their customers. When a new customer comes, they need a new VM. So when their customer signs up, they programmatically (via the API) make a call to us, create the VM, and they present the VM with their specific template that they've created and they present that to their customer and their client API with various functions like reboot/resize/snapshot/assign IPs, etc. Via the API, they can deploy a new customer and have the customer programmatically manage the resources of the VM. This is interesting because the customer might not know Cartika is in the background and yet the SaaS provider is able to deploy the product programmatically without touching our platform. Tamar: Tell us more about the virtual desktops offering you provide. How popular is it? Andrew: It's more popular than we thought. We're finding the usage is really high. The virtual desktop offering is a specialized virtualized machine offering and is presented and built fundamentally because of how Microsoft licensing works. It's more cost effective to build out environments specialized for that usage. You're able to find a much better access point as a virtual desktop instead of providing a virtual machine with Windows/SQL. We're finding people layering terminal services on top of them, using the product as a centralized workstation. Some customers run legacy applications that need to be offloaded to hosted cloud, such as CRMs like GoldMine or Quickbooks or applications like that. A lot of companies are offering cloud solutions and are limited in terms of the support. Our customers are able to run what they need on the cloud via terminal services and via virtual desktop vs. running a physical server in the office and having everyone connect that way. Tamar: I see that last year, you began offering real-time analytics for your hosting platform. Can you tell us more about that? Is it a proprietary service? Andrew: It's really cool. It's a New Relic integration. We integrated with them some time ago and are currently using it for our managed customers. (If you have a pure infrastructure only VM, you can install it yourself.) We're graphing the VM details for our customers in their client portal. We set up auto notifications for them so they are notified about CPU/RAM/disk partition issues or server slowness to be more proactive in terms of server maintenance. We offer upgrades to the enterprise platform for those who want it. Our roadmap will expand upon the offering, giving people the ability to get deep data analytics in their application to see bottlenecks and to see processes that can be optimized. It's also worth mentioning that we're going to make it available even for non-managed customers, featuring a single click install to install New Relic on their IaaS deployments. Tamar: I see you came from IBM as an account manager, which you mentioned earlier as well. What experiences have you brought over from your time at IBM to Cartika? Andrew: IBM back then was 15-20 years ahead in its thinking. It really opened my eyes to where the market and the world was going. To be honest, we haven't caught up to where IBM is going just yet. For example, on-demand computing was offered then--which is essentially the cloud (IaaS) everyone is offering today I say we're really progressive and they really showed me where the market was going and where the opportunity was. We still operate on the vision that IT as a service is kind of where the market is going to go; infrastructure or associated services or managed services, as well as backup and recovery, are all really the future of where things are going. That was the big takeaway from my past experience, where I got insights on where the market is going to go, how it's going to evolve, and yet, I still think we're 5-10 years away from reaching that point. Tamar: What inspired the name Cartika - what does it mean? Andrew: It means "paper" in Greek. But truthfully, it's inspired by a Mexican beer as well that I discovered when on a backpacking trip through South America while soul searching about my future. I decided to start Cartika as a result I was drinking a beer and had my thumb on the label and those were the letters that appeared on either side of my thumb. I actually do not remember the name of the beer, but, it was something like "cart" - insert thumb here - "ika" For a few years, I had that name on the back of my mind. When I decided to leave big corporate and to go off on my own, I registered the domain name and started up. Tamar: I love how Cartika is more on the cutting edge when it comes to services among the hosting companies I've spoken to. What evolutions do you see coming in the next few years and how do you plan to address these needs? Andrew: We used to sell everything bundled via server/VM, etc. We evolved by separating into the pure infrastructure and the associated services in a modular format. The next evolution is the continued automation of our management platform to the point where we can layer our management platform on infrastructure anywhere. We only manage backups and disaster recovery for FISMA/HIPAA but we'll be transposing services on other infrastructures soon, such as Amazon and Google - management as a service. These platforms are pure IaaS. They have a partner community layering value on top of the infrastructure so we're building a similar model. I can really see our management and our disaster recovery (DR) services being another layer on top of different infrastructures. Tamar: You mention proprietary disaster recovery and backup services that you offer. Please tell me a bit about this. Andrew: We went out and utilized the open sourced backend bacula.org to build our own product. We're like RedHat to Linux for this open source product. We built a whole suite of automation tools that let you backup and recover servers, databases, and VMs. It's a fully automated and integrated platform. For us, we have it integrated in our platform. As long as you have the proper management level and/or the backups and DR services on top of the IaaS, you can granularly restore whatever you want and set retentions as needed. The tool helps you resolve issues regarding backup failures and issues such as that as well. It's a pretty cool platform. Currently, we're focused on offering this tool for other service providers. In the near future, we're going to be focused on backing up and recovering your cloud. Whether you're on Amazon or your own private cloud, we'll provide it as a SaaS model or an on-premise hosted model. Right now, we charge by the number of VMs and servers being backed up; there is a cost per gig to backup data to our cloud or per device locally and there's no charge per gigabyte if you utilize your own storage devices. If you already have a server somewhere, you can back up all your data and there's no charge per GB. Tamar: What else would you like HostAdvice readers to know, if anything? Andrew: The world's changing and a straight infrastructure solution is difficult. You really need, whether you have your own IT staff or not, complimentary services to ensure 24/7 monitoring and assistance at times like 3AM when your staff isn't available. The real value in moving data and services to the cloud as a service model is associated services and support and that's where the really big boys drop the ball and that's what a lot of organizations are discovering now.