Jump to content

Jesse Howard

Customers
  • Posts

    10
  • Joined

Posts posted by Jesse Howard

  1. We actually believe Comcast was de-prioritizing UDP SIP/RTP for a customer (different voice platform) since by sending the traffic over a GRE tunnel between the customer router and our data center where they couldn't inspect it and it resolved their signaling and audio issues. Using SRTP didn't really solve anything because the headers are still available for identifying traffic type but the tunnel made a difference.

    Fairly certain that is illegal (continue to just say NO to internet fast lanes!) so good luck getting them to admit to it but something to consider when using ISPs who offer the same services on their own Network.
  2. The page here suggests this might be there already on-demand through a support request. Looks like the pre-requisite is you must have your shoutcast url handy and the location must end in mp3.

    "As per another ticket (UI-386), Darren and Karl A have noted that this is already supported for the most part: Darren: On a technical note, we can allow the user to enter any HTTP or HTTPS url that they wish, however it must end in .mp3 to invoke the shoutcast decoder"
  3. I only have a small proof of concept using the webhooks for updating call/presence events while using the channels API to initialize/periodically keep things in sync. Seems to work okay. The user discovery process is a bit of a pain because a customer may add a few new users and devices but until they fire off a new call there is nothing to kick off the discovery process automatically so you have to periodically check for changes on a schedule which seems so terrible to me these days with webhooks right there. I was hoping to use a webhook event based on account changes that I thought I heard mention of at kazoocon but haven't seen anything that triggers on user or device MACD operations.

    @Darren - any info on how the websockets API will look in terms of entity discovery and initializing presence display across an account or sub accounts?
  4. We have had some success using Nuance (actually used Ditech who got bought by Nuance) but our expectations are low - there is usually enough context to figure out what is being said. Nuance does offer services where you can send it back and for a charge have a person transcribe it for you but that gets muddy with billback etc.
  5. I am looking through the API documentation and under "realtime call control" it offers information about how to handle/manage call flows, not realtime manipulation of active calls that have been answered by a device as I would expect from the name. Am I missing some keyword in my searching or is this truly not available?

    Use case: with webhooks (soon to be websockets) I can get a screen pop of caller information and perform dips into a CRM on an inbound call to display account and recent case information. Ideally that popup also contains one-touch buttons to place the call on hold, park or transfer to the last rep etc. (similar to various existing SalesForce CTI integration apps).

    Ultimately the question is: am I spinning my wheels as an API to accomplish this doesn't exist (yet?), or is my vocabulary lacking the correct terms when searching?
×
×
  • Create New...