Twilio’s OpenVBX: Open Source Attendant

I downloaded OpenVBX, Twilio’s bendable, programmable cloud-based unified communications platform, tried out a few call control flows, and then drifted off into a reverie about telecom start-ups before the crash.

When the CLECs and ASPs first came on the scene in the 90s, they were offering hosted personal attendants (or assistants)—which was the term used before “Google Voice-like”—that allowed subscribers to configure find-me/follow me schedules for cell, home, and office numbers,  set up voicemail notifications, and craft simple IVR menus. They would often  throw in  speech rec, and support virtual presence through local phone numbers.

Maybe $30 per month, with a cap on minutes. These personal auto attendants were  tasty telecom appetizers and considering what was available from incumbents at the time, practically disruptive.

I looked through some old magazines articles to refresh my memories, and came across such players as Linx Communications and Webley and Wildfire Communication that were doing Google Voice platforms before there was Google. (There’s a link to Guy Kawasaki’s 1998 review of Linx at the end of this post.)

The big difference between then and now is that OpenVBX  makes it  much, much easier to build a serious IVR-auto attendant, one that can have deep hooks into your IT  infrastructure.

I like OpenVBX.  It is an open-source service creation environment in which the call switching  and UC part takes place off-stage, in the cloud. It is  well suited for small, tech savvy companies that have an existing, but limited phone system or none at all and run their business from employees’ mobile phones.

For them, OpenVBX will  be a unifying force, turning separate employee cell numbers into a single office number, and through the help of an auto attendant flow,  still reachable by the right DTMF key.  OpenVBX even supports a credible voicemail transcription service (based on my limited testing). But, sigh, no voicemail TUI plugin or applet as of yet—though with a little bit of programming someone in the OpenVBX community is sure to publish something.

One of the less than golden aspects from the golden age of personal assistants were their limited extendability. You could add features, but that often involved downloading a difficult to wrangle C++ SDK or else contracting for professional $services$.

This is not the case with OpenVBX’s sweet web design interface that has room for user coded PHP applets and relies on Twilio’s own TWIML (a VoXML-ish language) to define the call control and message handling:  roll your own or pull in code from the OpenVBX community.

OpenVBX graphical call flows

So the open source part of the OpenVBX is very attractive.   As has been mentioned by other reviewers, OpenVBX doesn’t support  SIP, the big open standards session protocol that underpins lots of UC platforms.  This is a less than desirable state of affairs as  SIP softphones (X-Lite, SIP Communicator) are excluded from the fun.

On the data side, OpenVBX  is promising.   By gluing  hooks for web services  to Twilio’s  UC capabilities, very  useful data apps—say, ordering, ticketing, sales tools — should be in the reach of a small business with a PHP-savvy web programmer.

Mini-gripe with the name: OpenVBX is not a virtual PBX; it is closer in functionality and spirit to the personal attendants I was waxing fondly about than to the hosted IP PBXs , the first true virtual PBXs,  which also came on the telecom scene in the 1990s.

Since I have Nick Carr on my mind lately (see my post from last week), I ask the question , does all this open telecom matter?

On one side, I’m not sure the open source community developers need to collectively re-invent ACDs, voicemail TUIs, and the zillions of other features that have been long established (and have been engineered for reliability) on small businesses PBXs and small-office-home-office “one box wonders” (see talkswitch and Zutlys) .

On the other side of the debate, OpenVBX is a malleable enough platform that can be designed to do lots of things that are not available from from the traditional vendors and service providers. Telecom does matter and those companies that master UC in the cloud will have competitive advantages.

Back to reality.

Overall, I think that OpenPBX will find its  mark, and their pay-as-you-go model ($1 per month for phone numbers, $.03 minute for calls) will be  appealing to cash strapped start-ups or small companies that don’t want to be bothered with installing and maintaining their own phone system.

Enhanced by Zemanta