API platform from Paris.
There are many API platforms out there for communications, and a lot of them are now offering WebRTC. As this is something I am going to divert my focus to in the coming months, I wanted to interview another such vendor. This time, coming from Paris.
I met Philippe Sultan in the last WebRTC Expo event in Paris. As Co-Founder and CTO of Apidaze his views were quite interesting for me.
What is Apidaze all about?
Our original motivation was to build an open-source cloud PBX and provide carrier services (DIDs, SIP trunking) for businesses of all sizes. Once we spent a year to build this, it became obvious for us to offer a REST API that would empower web developers to build their own cloud communication services. We can say that our goal now is to glue web applications and telco features. More in details, we provide our customers with everything they need to build up from simple communication apps to a fully featured voice & messaging system.
This “telco” material is completely exposed to web developers throughout three tools :
- REST API to place orders to our platform (e.g. place a call, send an SMS, create a SIP account, etc.)
- An easy to use XML scripting language that contain instructions defined by the customer and executed in real-time by our platform when processing calls (e.g. redirect a call, record a call, say some text, etc.)
- JavaScript API that allows developers to place calls from web browsers, using WebRTC or Flash
Our mission is simple: make web developers happy with telco by giving them the ability to develop telco applications without having to learn details. For that purpose, WebRTC is a huge opportunity to us, as it is now even changing what we call “telco applications”.
What differentiates you from the rest of the WebRTC API vendors?
Every company has its own history, we come from the open source telco world that is happily being disrupted by WebRTC. So I guess our background differentiates us from pure player APIs: our WebRTC API is just another point of entry to our telco platform. Another thing that makes us different to some others is non-WebRTC browsers. We want web developers to work with the widest range of browsers, this is why our JavaScript API sits on top of both WebRTC and Flash.
Why do you think there are so many WebRTC startups located in Paris? You’re like second after California in terms of number of companies.
WebRTC is a very appealing technology for web developers, and for developers in general. Apart from this very tech aspect, more and more people understand how they could use WebRTC and build new products out of it. I’d love to think the WebRTC meetups Bistri and us organize on a regular basis help spread the word here in Paris, but I believe the technology by itself is making its success among the developers located here, like any other place in the world.
What do you use for signaling?
We use our own signaling on top of WebSockets, mostly. No SIP nor XMPP here, just a WebSocket that carries raw JSON messages. This helps web developers easily build JavaScript event handlers, and process events transmitted at a fast pace. Remember our mission, we want to make web developers happy with telco, and one thing web developers are happy with are JavaScript events. And raw JSON seemed to us to go straight to the point for that simple purpose.
How is your backend architectured?
We use open-source software only (FreeSWITCH, Kamailio and Asterisk), and the great minds behind those tools quickly adopted WebRTC. Indeed, an SDP parser, a DTLS-SRTP + RTP stack coupled with ICE is all you need to build a PeerConnection and start a voice/video chat with other WebRTC endpoints, right? Well, those ingredients are present in those software packages.
Also, we’re not doing any direct peer to peer connection, as all calls being placed from WebRTC go through our WebRTC/SIP gateways.
Given the opportunity, what would you change in WebRTC?
I’m certainly not smart enough to answer that one, so I won’t. But I’ve been reading many posts saying how bad the choice of SDP was to exchange capabilities, and well, as a VoIP engineer / C programmer who works with open source software (not speaking as a web developer here), I must say that dealing with SDP was an easy path towards having an RTP instance, and therefore a PeerConnection. However, I must admit I made my opinion walking only in the dedicated VoIP field of WebRTC.
What’s next for Apidaze?
Many things going on. We need to add more features, and focus on mobile devices. We’ll keep on bridging the web with telcos with WebRTC! And getting more developers and customers to build great solutions with our API.
–
The interviews are intended to give different viewpoints than my own – you can read more WebRTC interviews.
Philippe, your approach is very similar to Twilio’s – cloud communication services with REST, XML-ish and client/browser APIs. Twilio is also heavily into SMS as well as voice, with some WebRTC today but not video yet. How will Apidaze differentiate itself from Twilio, and others in the area like Plivo and Tropo?
Hi Lawrence (and Tsahi)
As you mentioned, Apidaze is rather in the cloud communications API space much the same way as the companies you mentioned.
We believe that WebRTC is a disruptive technology in despite of the fact that no “killer app” has yet emerged, but will certainly do in the next months or year. There are great WebRTC developer APIs in this area, and Apidaze’s mission is not to create another one. Video is a critical application, but voice & SMS automation are still more critical.
The applications a developer can build with the Apidaze API are ready to use immediately and are not dependent on technological availability; most of WebRTC companies rely heavily on progress made by third parties. We prefer to think that our customers need to move forward with immediate solutions and adapt to their environment.
Tsahi mentioned after the WebRTC World Conference in Atlanta that he was expecting more innovation from WebRTC pure players on the use of the data channel and I totally agree with him. For the time being, most WebRTC startups have been exploiting the media channel; we use WebRTC as just another point of entry to our telco platform.
But we continue our R&D on WebRTC and will probably work on innovating solutions using video.