Posts Tagged ‘Cloud Computing’

What is “cloud computing”? Slashdot debates the question…

Friday, July 18th, 2008

No sooner had I pointed here to my article on IVR.tmcnet.com about pushing IVR into the cloud that contains this comment about the term “cloud computing”:

So What Is Cloud Computing, Anyway?
In recent months, it has become hard to escape the term cloud computing. Articles about the cloud are appearing everywhere. Conferences have been formed. Media Web sites have launched new cloud computing blogs and portals. Yet like any new term, the precise definition varies depending upon who you talk to.

than a colleague pointed me over to a discussion on Slashdot, “Multiple Experts Try Defining ‘Cloud Computing’“:

jg21 writes “Even though IBM’s Irving Wladawsky Berger reports a leading analyst as having said recently that ‘There is a clear consensus that there is no real consensus on what cloud computing is,’ here are no fewer than twenty attempts at a definition of the infrastructural paradigm shift that is sweeping across the Enterprise IT world — some of them really quite good. From the article: ‘Cloud computing is…the user-friendly version of grid computing.’ (Trevor Doerksen) and ‘Cloud computing really is accessing resources and services needed to perform functions with dynamically changing needs. An application or service developer requests access from the cloud rather than a specific endpoint or named resource.’ (Kevin Hartig)”

The links and discussion do make for interesting reading (in the Slashdot-kind-of-way) and point to the fact that the term is still evolving… and probably will for quite some time. I’ll still stand by my definition in my article:

Cloud computing is the ability to run your applications on a providers’ computing platform out in the network cloud.

Basically, you can run your apps on someone else’s system on their network. Rather than having to maintain your own servers, you can push that computing out into someone else’s network and make the system administration their problem. The platform you use can be some massively distributed and massively scalable network — you don’t actually care as long as it works.

In my article, I also go into a bit of my view of how we got to the place we’re at where the whole idea of “cloud computing” makes sense.

How do you define “cloud computing”?

P.S. The Joyent video I pointed to earlier also makes for entertaining viewing on this subject…

Technorati Tags: ,

Voice & Cloud Computing: “Pushing IVR Into The Cloud, Part 1: Making the Move”

Thursday, July 17th, 2008

clouds-kevindooley.jpgSo what are the steps of pushing your voice applications out “into the cloud”? Where do you begin?

As readers may sense by now, I have a fascination with network clouds (and cloud computing), and in the latest sign of that, I’ve kicked off a 4-part series over on the TMCnet IVR Community that we are sponsoring. In this first article in the series, titled “Pushing IVR Into The Cloud, Part 1: Why Make The Move?“, I talk about why you might want to consider moving your voice application out into “the cloud”. Here’s the intro:

With the world of Voice over IP (VoIP) comes an incredible number of options for how we configure our voice systems. Telephony can now be massively distributed over large networks. Remote workers or agents can be easily linked in to the rest of the office system. And many options exist to connect your VoIP telecom system to the Public Switched Telephone Network (PSTN).

In the midst of all of this is the challenge that application servers, such as those used for Interactive Voice Response (IVR), typically sit on a server somewhere in your network. And you need to make these application servers accessible to all the distributed parts of your network.

What if instead of running your IVR applications on a local server, you could run them “in the cloud?”

What if your IVR was out “in the network?” Accessible from all of your sites and from the PSTN? What if you could do this without adding to the load of your IT department?

In this four-part series, I’m going to explore what we are calling cloud computing and how you can push your IVR applications out into the network cloud. First, in this article, I’ll look at what cloud computing is and why people should think about it for IVR. Second, I’ll look at the security issues and what questions you need to be asking when considering a vendor for pushing your application into the cloud. After that I’ll discuss the key role that open standards play in allowing you to avoid vendor lock-in and I’ll conclude with steps you can take to make the move into the cloud.

The article continues on to talk about what cloud computing is, why you might consider it and why you might not consider it. Now the article focuses on “IVR” because… well.. it’s the IVR Community on TMCnet! However, you can pretty much go through that article and replace every instance of “IVR” with “voice application” or whatever term you want… the concept is the same.

As I note at the end, Part 2 will talk about the reliability/security aspects of pushing voice applications out into the cloud and the question: Can you trust the cloud to always be there for you?

Stay tuned for more… (and please feel free to let me know what you think as comments to this post.)


Photo credit: kevindooley on Flickr

Technorati Tags: , , , , , , ,

Can you trust the cloud (platform) to be there? (As Google’s App Engine goes down briefly…)

Wednesday, June 18th, 2008

googleappenginegoingdown.jpgCan you trust the cloud computing platform to be there when you need it? Is the platform reliable enough to always be available? If you are going to move your applications off of your own servers and “into the cloud”, can you trust that your customers will be able to get to those applications?

In light of Google App Engine’s downtime yesterday, people are asking that question again. TechCrunch’s article provided this cute little graphic on the right and said “Stuff like this tends to make developers nervous about adopting a new platform.

Downtime definitely does raise concerns, but let’s put yesterday’s Google App Engine outage in perspective - Google has been very clear that GAE is a preview release for developers to test applications. Nowhere have they said that you should run production apps on the GAE platform. Not yet anyway. It’s a place for developers to experiment - as I did with use GAE to store VoiceXML applications. Google has since explained the issue they found and at the moment everything seems to be fixed. That’s the reason they are running the beta test that they are… get a large number of people on the platform and see what issues are to be found. By the time they get Google App Engine out into production, you’ll have to think they’ll have sorted many of these issues out. But for right now GAE is definitely “beta”.

Now, when you look at cloud computing platforms to run your production applications, you do need to be asking questions like:

  • What kind of availability guarantees / Service Level Agreements (SLAs) does the platform vendor provide? Are there actual guarantees or is it simply “best effort”?
  • What kind of geographic redundancy is built into the underlying network? If there’s a natural disaster in one region will it wipe out all access to your applications?
  • What kind of network redundancy is built into the underlying network? Is the “cloud” all ultimately running on a single carrier’s connections to the Internet? (Bad idea?) Are there multiple carriers servicing each geographic location?
  • What kind of physical redundancy is built into the data centers? In the end you need electrical power, air conditioning and physical security to ensure that the actual servers upon which your apps are running continue to run. What is the “cloud” platform built on?
  • What kind of monitoring does the vendor perform? Is there a 24×7x365 Network Operations Center (NOC) that is monitoring the health of the network and all applications?
  • What kind of scalability is in the cloud computing platform? Can it grow as your application grows? Can it handle extreme spikes of traffic?
  • What kind of security, both network and physical, is part of the computing platform? What steps does the vendor take to ensure the network is protected from attackers?
  • Finally, what will the vendor do if there is downtime? How quickly will you be contacted? How quickly will the vendor respond? Will the downtime be reflected in your bill?

Here at Voxeo we’re obviously huge believers in moving voice applications into the “cloud”. We’ve been helping customers do that since 1999 and our hosted Prophecy IVR Platform is one of the largest VoiceXML hosting platforms in the world. We have a rock-solid, ultra-reliable, secure, distributed platform for running your voice applications. To our knowledge, we’re the only ones in our industry to back up that belief with a 100% uptime guarantee… which means we pay you if we have any downtime. (Subject to certain conditions outlined on the page.)

We welcome questions like the ones above. We’re glad to talk about our platform for voice applications. (Contact us if you want more info.)

You can push your applications out into the “cloud”… there are many happy users of cloud-computing services from Amazon, Joyent, our own platform and many others. But you do have to ask questions like these and understand whether or not the application platform will be there when you need it!

P.S. And for those who absolutely want an on-premise product, a few years ago we made our software also available as a premise product… we also support some fascinating hybrid models I’ll write about in the future…

Voxeo talks at Communications Developer Conference in September in LA

Monday, June 16th, 2008

commdeveloperconference2008.jpgWe’re very pleased to note that we’ll be part of three presentations coming up at the Communications Developer Conference (formerly “VoIP Developer Conference”) co-located with ITEXPO West 2008, September 16-18, 2008.

As shown in the Communications Developer Track 1 schedule, RJ Auburn will be up first on Tuesday, September 16, from 9:00-9:45am as part of a panel on “The State of SIP Application Development“. Here’s the abstract:

As SIP continues to be implemented in enterprise and consumer environments, applications are being developed that further leverage the IP capabilities of the clients. Two of these capabilities are location and presence — where I am, and what my availability status is. This presentation will address the various protocols and standards involved in presence and location, and how these apply in a SIP environment. It explains how SIP both enables complex applications that integrate unified communications with presence and location, but also some of the challenges that it presents.

Next, on Thursday morning, September 18th, Jose de Castro and I are back to back. The day actually starts off at 8:15am with what should be a fun talk on “Voice/Data Mashups” that includes Thomas Howe. After that, Jose’s on at 9:15-10:00am with “AJAX’s Impact on Telecom“:

AJAX has completely changed how developers approach the Web. A lesser-known and surprising phenomenon is AJAX’s impact on telecom and open telephony standards. This session will give the audience a brief history of speech technologies and how XML fits into the speech technology ecosystem. Jose will demonstrate how developers can apply their existing AJAX skills to build dynamic telephone applications, and why this development paradigm is ideal for telecom applications.

I (Dan York) am then following from 10:15-11:00am with “Developing Voice Applications in the Cloud“:

Today the industry is buzzing about cloud computing and pushing applications out into the network cloud. Google has brought out their AppEngine platform while Amazon offers their EC2 and S3 services and Microsoft and others prepare their own offerings. In this session, attendees will learn about what is involved with developing voice applications on cloud computing platforms. What options are out there? What do you need to look for in a platform? How can you get started? The session will include demonstrations and prepare the attendees to return home and get started.

We’ll also have a booth there and have now also become a Platinum Sponsor of the event as well. We’re a huge fan of developer-centered conferences (hence our strong past support of ETel and then eComm) and we’re looking forward to seeing how this one goes co-located with the large ITEXPO show. If the Communications Developer Conference is not yet on your agenda for the fall, we’d definitely encourage you to check it out!

Technorati Tags: , , , , , , , , , , ,

Launching your own *platform* for voice applications using Voxeo’s computing cloud

Tuesday, May 27th, 2008

You’ve seen our Evolution developer portal, right? It’s home to over 30,000 developers who have collectively built over 55,000 voice applications. It’s used by everyone from some of the largest companies out there running their customer service applications on our commercial Prophecy Hosting service to tiny startups using just our free developer accounts to look at how they can add voice to their new Web 2.0 site. On the back end, it’s a massively-scalable redundant cloud-computing infrastructure that executes applications and provides connectivity to the Internet, the traditional phone network (PSTN) and the ever-evolving SIP infrastructure, plus handles all the billing and management issues.

Now imagine this:

What if you could have your OWN custom-branded version of our Evolution site?

What if you could have a voice application platform that you made available to your customers so that they could build voice applications?

What if you were:

  • a carrier/service provider looking to provide additional services to customers such as hosted IVR that could earn you additional revenue?

  • a vendor of telecommunications equipment without a hosted IVR/application offering looking to quickly add such an offering?
  • a large enterprise looking to enable your internal business units to easily and quickly create voice applications without necessarily needing to engage with IT or professional services groups?

Now let’s take it one step further and imagine this:

What if you could do all of this without having to provide/run/maintain the infrastructure?

No servers to provision. No infrastructure to build. No software to upgrade. No system administration. No connections to the PSTN to worry about.

Now stop imagining… that’s exactly what we announced today: availability over our IVR and SIP hosting infrastructure as a completely re-brandable Platform-as-a-Service (PaaS) offering. Now, you, too, can have your own branded voice/IVR/SIP application platform. You can sign up customers and have them run (and develop) their own VoiceXML/CCXML/SIP applications. If you want, you can charge them and bill them by the minute.

All by way of your own completely rebranded web portal. No need to mention Voxeo - in fact, your customers never need to know. It’s your service offering. It’s your platform. We just provide the cloud behind it all - and remove all the infrastructure worries. Need it to grow more and run more applications? Need more inbound phone numbers? Need to make more outbound calls? You don’t worry about that… our cloud adjusts automagically to handle your needs. We provide the IVR virtualization or SIP virtualization layer. You just get to focus on building applications.

From a graphical point of view, it basically looks like this[1]:

voxeopaasoffering-5.jpg

Now the obvious concerns when contemplating a move like this include: How reliable is the platform? How secure is the platform? To answer, I’ll point to these parts of the news release:

Voxeo runs four fully redundant, proven, and patented hosting facilities from geographically diverse locations in the US and Europe. These facilities house over 800 servers today, each capable of processing over 200 concurrent phone calls. Through Voxeo’s patented platform architecture, this cloud-based, multi-server infrastructure appears to customers as a single virtualized, multi-tenant IVR system with nearly unlimited scalability. Thanks to Voxeo’s network diversity customers know their IVR failover and scalability requirements are comprehensively addressed. The hosted offering is completely managed and controlled using Voxeo’s brandable online customer and developer web portal, which includes Voxeo Designer, a flexible Visio-like visual telephony application development tool. Enterprise applications and telephony resources interact with the platform via standards such as VoiceXML, CCXML, HTTP, SOAP and SIP.

Voxeo has invested over $50m in the delivery of its IVR hosting network and continuously adds new capacity as customer demand increases. Voxeo’s platform includes comprehensive call control features to make, take, route, and connect calls; as well as robust voice media features to play, record, conference, recognize, and synthesize speech in 12 different languages. Voxeo will open a fifth hosting facility in Europe this year and anticipates launching facilities in Asia and other markets in 2009.

Obviously for customers serious about our PaaS offering we’ll go into more detail. We are also one of the only ones in the industry to offer a 100% uptime guarantee (or we pay you) and our PaaS offering can be configured in such a way to fall under this SLA.

I’d note that this Platform-as-a-Service offering isn’t just some nice theory… it’s out there being used today by a very large customer as an OEM platform. It’s very real.

Interested in learning more? Would you like to offer a complete (and branded) voice application platform to your customers or users? If so, drop us an email to sales@voxeo.com or use that phone thingamajig to call +1-407-418-1800.

[1] Although to be completely accurate, all those arrows should have an arrowhead on both ends because the information flows in both directions.

Technorati Tags: , , , , , , , , , , , ,

Joyent video: What is Cloud Computing?

Tuesday, May 27th, 2008

In the weeks and months ahead you’ll hear us talk more and more about “cloud computing” as that term becomes increasingly mainstream (as I did in our Speaking of Standards blog last week). To us, it’s not at all a new concept as it’s what we have been doing with our platform since we launched way back in 1999.

With our hosted platform, we give you a massive computing cloud upon which you can run your voice applications. You just create the application and let our platform (i.e. our “cloud”) worry about scaling, redundancy, computing power, etc. (Our Quick Start Guide gives an overview.) There’s no need to worry about how many “ports” you have or how many server instances or what type of CPUs you are using. All of that fades away. You develop your voice applications (IVR apps or SIP apps) and run them… and our cloud takes care of all the infrastructure.

Of course, if we are doing to talk about the term “cloud computing”, it helps to provide some definition. On this note, I have to really hand it to the folks over at Joyent who compiled this excellent video of responses to the question “What is cloud computing?” from attendees at the recent Web 2.0 conference. Links to the people you’ll see are available on Joyent’s weblog (”Joyeur”) but courtesy of YouTube here is the video:

While Joyent is focused on providing cloud computing for rich web applications, the responses in the video are equally true for the voice applications that we support on our cloud computing platform. (And, like Joyent, we are strong believers that any cloud computing needs to be based on open standards to avoid vendor lock-in.)

Ultimately, what we now call “cloud computing” is the fulfillment of the dreams many of us have had since the early days of computing… creating a massive “cloud”, comprised itself of numerous smaller clouds, which let your applications “just work” and allows you to focus on building killer applications and not having to worry about the infrastructure to run those applications.

It’s a pretty amazing time right now - and we’re thrilled to be part of it all!

Technorati Tags: , ,