»

v01c3.com - @ChrisMatthieu's Voice Blog

Chris Matthieu is a telecom executive and next-generation communications hacker extraordinaire sharing his first-hand voice experiences with the world.

Home Archive RSS Ask

WebRTC Engines

With all of the justified excitement around WebRTC lately, several new JavaScript WebRTC frameworks have emerged!

The first WebRTC Node.JS module to hit the market a little over 6 months ago was WebRTC.io. At the time of this article, it has 222 star gazers and 43 forks on GitHub.  It supports GetUserMedia and PeerConnections on Chrome browsers and it has a great conferencing demo.  You should definitely checkout this library.

Newcomers to the developer race include: Holla, PeerJS, and EasyRTC. At the time of this article, none of these frameworks support FireFox; however, my bet is on Holla to have it integrated first.  All three of these WebRTC engines support GetUserMedia and PeerConnections on Chrome and come with simple demos to get developers started building voice/video apps today.

In addition to GetUserMedia and PeerConnections, Holla also supports P2P calls for both placing and receiving calls as well as handlng chat and presence.  Note: Holla is the WebRTC module used in Twelephone.

This is a fast moving space and it’s great to have options. Stay tuned…

I recently joined @mjgraves on the @VoIPUsers Conference regarding my new HTML5 WebRTC project @Twelephone!  We covered@Twelephone’s service, roadmap, architecture, and WebRTC in general.  I especially enjoyed the reunion with @Steely_Glintfrom the Voxeo Phono days. 

The VoIP Users Conference is a weekly live discussion about VoIP, SIP, Asterisk and all kinds of telephony-related topics. The conference has been running for over four years and consists of the leaders in the VoIP industry.

Here’s my WAV/MP3 to text dictation POC!

Services like Google Voice, Tropo, Dragon, and PhoneTag have offered dictation services for a while now; however, the quality hasn’t been exceptional.  Google and Nuance (with their Dragon product) have probably invested the most money into R&D of natural language speech-to-text (dictation) technology.  The problem is that all of this technology is closed until recently.

Google introduced a microphone icon for use in form fields for users using the Chrome web browser.  There have been many reported hacks to leveraging this mic technique and the undocumented Google Speech API for other purposes.  I recently stumbled on @MaxOgden’s Stenographer Node.JS module which got me thinking…

We could leverage this technique to stream WAV or MP3 audio files into Google’s Speech API in small chunks and have Google stream back the text to our app!  If the audio you pipe in is longer than 10 seconds it will automatically get chunked and processed 10 seconds at a time. Google’s servers return errors if you upload large files but 5-10 second clips seem to work okay.

This project was just an experiment but it’s interesting what you can do with Node.JS and speech these days.  Here’s my code: https://github.com/chrismatthieu/dictation-api

Just submitted my new startup, Twelephone, to the AWS Startup Challenge!

The telephone was invented more than 100 years ago. Don’t you think it’s time for a new telephone that works the way you do?

Introducing Twelephone - The world’s first HTML5 (WebRTC) voice and video phone that integrates with Twitter!

Excited to see that my @Nodester acquisition by @AppFog made the local newspaper!

Excited to see that my @Nodester acquisition by @AppFog made the local newspaper!

Making great progress on the Twelephone project! It now supports Twitter OAuth, DM/Tweets, presence, and call presentation with Accept and Reject options!

Another Startup Acquired! What’s Next?

As you have probably already seen on TechCrunch, GigaOm, VentureBeat, eweek, and DZone, I have sold my open source PaaS startup Nodester to AppFog, a leading polyglot PaaS service provider!

If you remember back two years ago, I also sold my cloud communications startup Teleku to Voxeo.  Ten years prior to the Teleku acquisition, I built and sold another company called Digital Voice Technologies which build and managed IVR (Interactive Voice Response) services for companies like Disney, Coca-Cola, and Microsoft.

All of three of these companies were bootstrapped subsidiaries owned and operated by GetVocal, inc, my AZ-based technology company, which made the exits easier.  Reflecting back on these deals has generated a few questions and thoughts:

  • Does this make me a serial entrepreneur? 
  • Does this make GetVocal an incubator?
  • What’s next?

I enjoy the excitement around a new idea and opportunity.  I love the branding side of the business so much that I have hundreds of web domains and can’t write a single line of code until the project has a name. I love solving new problems that have not been tackled by other companies or projects.  I love the notion that my idea has the potential of changing the world.  I love the excitement surrounding launch day. I love meeting new people throughout the life of the project. I love coding but equally move the marketing and sales side of the business. Finally, I love negotiating the exit deal and watching my project take on a new life under someone else’s vision. 

This process just seems to continue repeating itself.  

If you visit GetVocal’s website, you’ll notice that I have another project already underway.  In fact, it’s be underway for several years now.  This project is called Surf-By-Tel and it’s more of a humanitarian / labor of love project enabling visually impaired people to surf the web with their voice.  I love seeing a visually impaired person’s face light up when they first learn that they can surf the web with their telephone of VoIP application.  Check out this video testimonial from a user of my Surf-By-Tel voice browser platform:

It’s hard to say what’s next. I have a few ideas that I am experimenting with but I always fall back to something Don Valentine from Sequoia Capital once told me, “It takes just as much work to start a million dollar business as it does a billion dollar business.”  This means that my next startup will need to be even bigger yet so stay tuned…

This is an alpha demo of my HTML5 WebRTC-powered Twelephone platform.

Node.JS HTML5 WebRTC Websockets Voice/Video VoIP (did i miss any buzzwords?)

My CouchDB Notes

CREATE COUCHDB DATABASE

curl -X PUT http://127.0.0.1:5984/books

WRITE RECORD

curl -X PUT http://127.0.0.1:5984/books/1 -d \

”{

"_id":"1", 

"title":"couchdb the definitive guide"

}”

READ RECORD

curl -X GET http://127.0.0.1:5984/books/1

DELETE RECORD

curl -X DELETE http://127.0.0.1:5984/books/1?rev=1-6e488095a5bf295b5880ce74846ca7bf

UPDATE RECORD

curl -X PUT http://127.0.0.1:5984/books/1 -d \

”{

"_id":"1", 

"_rev":"3-129ee86f127191d76c5560f10f06d082",

"title":"my book"

}”

MAP:

function(doc) {

  if (doc.title){

    emit(doc.title);

  }

}

Save as defaults / titles

GET VIEW

curl -X GET http://127.0.0.1:5984/books/_design/default/_view/titles 

GET VIEW with Document

curl -X GET http://127.0.0.1:5984/books/_design/default/_view/titles -G -d include_docs=true