That is a really good news! I have been using Twilio for a client of mine offering callback numbers for one of their product. I found out about Plivo last March at PyCon in the expo hall, and their price where 50 to 100% cheaper, while providing what I considered a similar service. We deployed with Twilio, but our next milestone includes switching to Plivo. Best of luck to them!
Hopefully the folks at Twilio are following this thread.
For all the awesome things about Twilio, its pricing at serious volume is very shitty. I was recently talking to a start-up that makes apps for call centers. He blamed the super high price relative to their competitor on Twilio's pricing. The start-up's competitor uses Plivo.
May be Twilio believes they can survive without taking part in price wars. We'll see.
Familiar situation, but in the longer run, the minute pricing from resellers is less important. What Plivo is doing here is very smart - they are leaving the lower margin pricing games to carriers and focusing on the true value-add, which is the business logic stack.
Twilio is following these moves too with their SIP-out product. I'm sure it's only a matter of time before the support receiving SIP calls and registering endpoints.
For smaller customers that do not have the buying power necessary to secure proper redundancy, both Plivo and Twilio offer minutes. Plivo's emphasis on more advanced telephony management is definitely a good differentiation from Twilio, who is focusing more on customers that don't know much about telco.
Remember, if you are routing 10s of millions of minutes, you likely have at least a few people focused on supporting it and revenues to underwrite carrier relationships.
DISCLAIMER: I am not in any way affiliated with Plivo, Twilio or any other provider.
Twilio, Plivo, Tropo/Voxeo, etc. are reselling minutes at a given rate that are aggregated from a basket of carriers. You get a "blended" rate based on their ability to optimize the routes (least cost routing or LCR), carrier quality/redundancy needs, and margin the reseller is trying to make.
For smaller companies, individuals, these blended rates is absolutely the way to go - less complexity, less variability in expenses based on traffic, fewer things to screw up.
For a larger company that already has the carrier relationships and staff to manage them, these reseller models become prohibitively expensive.
If you come to an IT Director managing telco for a large company and tell him/her that it will be $0.02 per minute in 60 second increments (Twilio's list rate) to make outbound calls, you will get laughed out of the room.
Of course, Twilio, Plivo and others offer volume rates, but they still need to make margin. If you are big enough to have direct relationships and/or buying multiple services from carriers (such as minutes and bandwidth), you get better rates.
Plivo is therefore very smart in going for infrastructure play. Twilio is also doing this with their SIP offering, but at a somewhat different angle.
OK. So the argument, if I might restate it to make sure I follow, is that Plivo knows the big companies have the relationships and clout to not need the price abstraction benefits they provide. So Plivo is letting them plug their own carrier relationships into its software so that they at least control the telco "business logic" for those companies, even if they're not able to get their aggregate, steeper price for the minutes themselves.
Here is a story from using the Plivo APIs, that convinced me that these guys are for enterprises and serious businesses.
We were looking at different telephony APIs to build a highly concurrent calling system. With Twilio, we could only send one phone number in a single http call. That was a bummer. With Plivo, we could send as many phone numbers as a http POST can handle, and let their APIs handle the high number of calls per second. We were pleasantly surprised that they do so much of the heavy lifting.
We are currently testing with 50 calls per second(cps) and Plivo just eats it up! We'll crank it up more later this month. I don't know what is the max cps Plivo can handle though.
I doubt anyone knows max cps numbers on any platform. It's usually a question of adding more servers until the software breaks, but there are lots of strategies. It always comes back to design.
Great news - Plivo is awesome my company is currently working on a VOIP system these guys are powering us. The call quality is fantastic compared to Twilo
There's a need for us carrier types to agree on a single-packet auth model. Basically, a way for both sides to agree on an an acceptable nonce, so HTTP Digest can be used without having to get a new nonce from the server.
Absolutely. With IPv4 becoming scarce this is surely needed. However, we are working with select few carriers to see if we can have this done the right way!
I wish there was something like openvbx.com for Plivo, or better yet, a carrier agnostic telephony client platform that could showcase the full functionality of the platform.
Yes. I built something on top of OpenVBX a while back: https://github.com/andreioprisan/igroups
I wish there was a universal web-based VBX that would work with a number of providers, either self-hosted or hosted by the platform provider themselves.
Our system is designed to run on any of the modern cloud telephony APIs: Twilio, Plivo, TelAPI, even your own RestComm stack. I don't know of anything OS though.
