Custom Messaging to Individual Finesse Agents

[claudioillanes]

Hi  All,

I want to get some guidance on the suggested implementation for the following. I've got a global .NET service that needs to relay messaging to about 5000 individual Finesse 10.5 Agents. I believe there are many ways to do this, but wanted to know what everyone thought would  be the best (and supported) way to relay these messages to an individual agent?

Should my .NET service connect with the OpenHub 1/2 concept and just relay text messages "user.domain.com" message topics?

Or would I want to add RabbitMQ to my .NET service, and build a gadget that uses a RabbitMQ Client to receive messages?

Or should I have the .NET service make a web service call directly to a gadget, that's somehow listening on a port?

Or should I tie in Socket.IO to my server and gadget? I basically want my service to be able to tell 'agent 11111 to 'Go Ready', or 'Answer Call'. I'm wondering if anyone has had success with this kind of unidirectional, or bidirectional messaging setup in the past, and successfully handled any load/performance issues. Also, has anyone had success relaying this broadcast messaging to agent teams? or Skill Groups?

Cisco BU, what's the recommended/supported setup for this?

Thanks!

Claudio

[dekwan]

Hi,

I do not have any experience building this type of gadget so I cannot really provide any thoughts on what is the best approach.

Our development partner 2Ring has built a gadget that contains what you need: Supervisor to Agents broadcast messaging in Cisco Finesse | Miroslav Moravek | LinkedIn

Thanx,

Denise

[claudioillanes]

Hi Denise et all,

So if I wanted to have the ability to send one  'custom' message to one specific agent (out of 5000) from a custom server, could anyone suggest the recommended way to relay this message to that agent? Do I tie into OpenHub 2 and have each agent subscribe to their own feed? I also looked at sending this through Notification Services, but it seems that I'd have to login each agent, or at least provide passwords when asking the system to 'place an agent ready'. Thoughts from anyone out there? Is it safe to just use websockets or socketio outside of the Finesse framework?

Thank you,

Claudio

[SEAN NILSEN]

Hey Claudio - I see this is an old thread but was looking for something else and saw this.

Socket.IO is definitely the way to do this. It requires you to build a server-side component to manage the socket connections but is very do-able on a per-agent, per-team, per-whatever-logical-split basis.

If you're still poking at this, give me a shout.