Share This Post:

Skype line for predictive dialer agent

Does the skype instance run locally on the machine with the “Agent Dialer” app on it (Client computer), or does the skype instance run from the machine with the “Gateway” app on it (Server computer) ? With the eval version I’ve got, I seem to be getting the impression that the “Gateway” app is the piece thats doing the actual invoke of Skype?

The gateway software on the server computer does all the calling. The Skype instances are controlled by the gateway software.

The “Agent Dialer” app on the client computer (agent’s computer) is called RemoveAgent.

If skype runs centrally from the machine where the Gateway app is running, and you can only run (1) Instance of Skype per machine, how do you have more than 1 agent?

You can run multiple Skype instances on a single PC.

Could this be configured so that the Skype instance is local to the “RemoteAgent” app on the client computer?

You can run the whole software (gateway and AgentDialer) on an agent’s computer. You can run single line or multiple lines. This way each agent’s calling is independent of other agents. But if you have multiple agents, and you want them to share a single call list, then you need a server computer for all the dialing (running skype instances), and have the agent’s computer just to receive the screen popup and audio redirection.

This doesnt really make too much sence, if the Skype instance needs to be running locally to the server computer (and not the agent) How do you get more than 1 talkpath to accomidate more than 1 agent in the enterprise edition for the people that are not sitting infront of the server computer running the gateway app?

The Skype audio is redirected to an agent’s computer. This is the only way to have multiple agents sharing multiple lines.

If you do want dialing from an agent’s computer, then simply install AgentDialer on each agent’s computer.

Perhaps Im getting myself a little confused here… My understanding is that Skype does all of the dialing thru their service, and that the “server” computer running the “gateway” app ties into its local instances of skype to pass calls up to the skype cloud?

If this is correct, how do the agent computers get a voicepath between the agent computer & the “server” computer running the “gateway” app that is framing the calls to skype ?

Yes, you are correct.

Here might be where the confusion is. Agent Dialer has two operating mode: predictive dialing and semi-automatic dialing. For semi-automatic dialing, an agent has to operate the line (Skype instance), such as to dial next number, hangup, play audio, record conversation, etc. With semi-automatic mode, each computer can only operate a single line, and thus only supports 1 agent.

However, in predictive dialing mode, calls are made from the server computer, and whenever a call is answered by a live human, it is immediately transferred to an available agent. The agent will see a screen popup window, detailing the called information, and he can talk to the person on the line using a computer headset (using computer’s microphone and speaker). In Enterprise Edition, you can dial out using multiple lines, and transfer live answered calls to multiple agents.

The talkpath from the server computer to a customer is the Skype instance, and the talkpath from the server computer to the agent’s computer is regular audio redirection. The software that controls the audio on the client side is the RemoveAgent software.

If someone could explain the concept of “Regular audio redirection” to me,for how the calls setup & torndown between the “Gateway” (Server) computer and the “Agent Dialer” (Client) computer works -That would be great…

When you use Skype to make a call, you use the computer speaker and microphone to talk to the other party. The microphone sends the audio stream to Skype and the speaker gets the audio stream from Skype. In a way, you are using the computer as a telephone handset. Now, when Voicent predictive dialer makes a call and it is answered by a live human, Voicent gateway, the server software, sends/redirects the audio stream to the client computer, so the agent can talk and listen on his computer instead of the server computer.

Ok, so if I have remote users that I want using the centralized database/list so theyre not double calling people on the same list that other agents are, what firewall constraints need to be made for this so they can get audio & signalling as expected?

For predictive dialing, it is strongly recommended to have the agent’s computer sits on the same local area network (LAN) as the server computer. If you have pretty good server and bandwidth, you can have agents work on remote locations. Since agent needs to login to the server in order to receive screen popup and audio redirection, you need to open up the port on the firewall from the server side. You do not need to change anything on the client side.

For more details, please select from AgentDialer’s main menu > Help > Predictive Dialer Setup Guide.

Just as a suggestion, this would work ALOT smoother if the skype instance were invoked locally on the agent computer (for both processing as well as network resources) and the server computer just passed it numbers from a central campaign database & kept track of the administrative agent data from all of the clients…

Yes, absolutely. And you CAN do that with Voicent software.

What you need to do is install Voicent Gateway on each of your agent’s computer, and have a centralized computer to run the AgentDialer spreadsheet as a client to all the available gateway computers. The AgentDialer will distribute calls to all the available gateway computers. To setup the remote gateways for AgentDialer, select from the program main menu > Predictive > Dialing Hosts.

Please note you need Enterprise Edition of the AgentDialer for each agent’s computer. An enterprise edition allows remote connection. Also, since the server (dialing computer) is on an agent’s computer, you need to open up the firewall on the agent’s side, since now the server is actually resides on the agent’s computer.

The advantage of this setup compared with the centralized sertup is obviously the scalability of the system. You virtually have no limitation on how many distributed gateways you can have for predictive dialing. The disadvantage is that the agent is mostly limited to receiving calls from his or her own computer. Because the moment you want an agent to receive audio redirection from a different computer, then you have to setup the whole thing like a centralized setup.

Overall, Voicent AgentDialer has flexible setups. The best setup depends on your campaign, number of lines, internet bandwidth, number of agents, agent location, and few other factors. Just experiment with the setups, you will find an optimal solution.

Share This Post:

This entry was posted in Predictive Dialer. Bookmark the permalink.