6 thoughts on “Introducing Cloud APIs for jambonz, Dave Horton”

  1. Thank you Dave for a truly impressive demo:

    1) What are some of the charging models you’re thinking about for Cloud APIs for jambonz?
    2) Focusing on voice makes perfect sense, you can not do everything at once. But is it video or messaging next?
    3) I love the analogy of: spilling the parts on the floor and saying, ‘isn’t it great?’ It’s always a challenge because the project is modular, you enable people to have complete control over their platform’s roadmap using your project, so they can operate at internet speed not telecom speed. As Cloud APIs for jambonz evolves, do you see it enabling people to play with some parts within the Cloud API service?

    1. 1) I’m starting with a very simple model — charging for application processing capacity, measured by how many simultaneous calls someone wants to support. On the plus side, this is a pretty well-understood metric in the telco space and, frankly, simple to implement. It probably won’t suit some customers uses cases (particularly those with bursty traffic), so its possible I will end up somewhere different, but to start with I think this is a sensible model to put out there and then start to get feedback on how to best meter the service in a way that supports customers different value propositions.

      2) Messaging is in the works, and I already have some providers saying they will contribute the messaging APIs for their networks, which is great. I do see video support coming down the road, but I want to build off a solid base of successful voice deployments first so that the brand becomes known for quality implementations. Then when I add video people will positive expectation that it’s been done right and they’ll want to try it out. But I am looking forward to working on video!

      3) Yes, definitely. Some of the information I want to get out there soon is an on-boarding tech talk for folks that want to contribute to the core of jambonz, which is pretty slim and trim in terms of lines of code (the last time I did a count on the core stuff it was less than 10K loc) so its going to be quite easy for people to add new commands to the json webhook “language”. I’m keen to get that going because I think its extensibility will be a big part of its success and something the open-source model offers its customers that proprietary CPaaSs do not.

Comments are closed.