## vCon "Bof" {#vcon-bof} * Not an offical WG, as the charter not approved yet. However,we expect an approval in 30 days with our current charter. We just can't make decisions. ## JDR Requirements {#jdr-requirements} * Overview of contact center technologies * Compliance issues * Quality Management * Speech Analytics * What was spoken * Angry customers, etc. ### Two flows: {#two-flows} * Real time or post call recording transfer, including call meta data and call events - needs to be shared with consumer * Desktop recording - not connected but also required and do a post call/upload overnight to supporting applications ### Meta Data Fields: {#meta-data-fields} * Interaction types * Interaction ID * File Type * Media meta data: bitrate, channels * Start time * Duration * Direction of call * Partipant details * PII and PCI Redaction * Skill * Campaign * Transferred Indicator * Conference Indicattor * Count transfers, conferences, holds * Hangup parth * Disposition * Dialing list ## Emergency Call vCON Use Case - Brian Rosen {#emergency-call-vcon-use-case---brian-rosen} ### Emergency Calls are Citizen to Authority {#emergency-calls-are-citizen-to-authority} * regulated carriers 3 digit dial code (911) IETF ecrit defined standards for emergency calls, routing of call based on location. Participants in an emergency call * caller * call taker * dispatcher * responders * all calls logged for legal purposes and QA by logging service * Privacy restrictions otherwise in force are waived 3 most important things about emergency calls:location Location is used to get to the right PSAP and responder Start of the call * qs for calls at PSAPS * You can get a busy, really want to avoid * True skill based routing is rare * Media is negotiated: audio, video, text Conversations * two way call always starts * Call is transferred to a dispatcher * Transfer is attended * Transfer to responder is rare, but same as attended * Data transfers usig standardized data is used to communicate incident data from call taker to dispatcher to responder Logging * all is logged, we use SIP REC, playback with RTSP * All calls have a globally unique incident identifier JDR: What is the unique ID requirement and what is the lifecycle? BR: We do it after the fact, but it depends on the actual use case. JDR: Wouldn't you want to keep this? BR: Depends when you make the vCon Persons * We track people with an incident, including non-victims and bystanders Operations on Conversation Data * Limited use of automated transcriptions, but changing * Logging is primarily for legal proceedings. JDR: Name the examples of where this goes between vendors BR: First, vendor lock-in. Secondly, in nextgen 911 vcons are sent during the incident. JDR: Also interesting in the failure network configuring use cases. ## Rohan Mahy, vCon for MIMI IM {#rohan-mahy-vcon-for-mimi-im} ### Three Use Cases {#three-use-cases} * Export conversation history from my old client to new device * Group chat with substantive discussion, then add some users. Allow an admin to "catch them up" by sharing message history from some relevant point in the past * Back up my messages New elements From the MIMI content draft * Under the parties list, imUri * Dialog : array of individual dialog * messageID: from mimi * parties: who is active in the MLS room at this moment * originator: party ID * In reply to: * replaces Content format for MIMI does not have these fields, so have to do that. RM: If they wanted to take this ## Dan {#dan} ### Learnings {#learnings} A few things we added: * created\_at * updated\_at GAP analysis, of use cases and requirements IDs Object signatures, how many variations of the same thing are we signing Give vCon a try ## Conclusions {#conclusions} JDR: Lots of metadata. We should have a baseline format, with super common stuff Cody: We are using vCons ourselves, looking at the extensions Cullen: We need an abstraction layer to define a core functional layer Brian: Maybe we actually want two layers, one well known, then one that varies per application.