vicidial.org

Hosted: nextsalesllc.vicihost.com | VERSION: 2.14-829a BUILD: 210911-1958 |

I think I mentioned the correct criteria above for making a post... I'm really just looking for someone to give me the breakdown of exactly how the predictive dialing method works in vicidial. I found a post in github that detailed it in vicidial.org/docs/predictive.txt document but the post was from a decade ago(so its probably been updated since then) and I'm assuming since then the document has been scrapped from the site since i cant find it? It detailed getting an average of agents waiting every 15 seconds as well as total leads, agents in calls average cust time etc. and i wanted to get that formula and process, wherever it may be so i can essentially replicate it in a python script that can calculate it from my machine and send api calls to change the dial level. Turning on adapt_avg within vici itself often overloads the servers, slows things down, etc. and usually just breaks things since we're dialing at a really high volume. Thanks!

check out my article about that:

https://cyburdial.net/vicidials-predict ... -revealed/

carpenox wrote:check out my article about that:

https://cyburdial.net/vicidials-predict ... -revealed/

Yes I saw that article, It explains how it works but not the criteria for how the dial level is calculated, i was looking more along of the lines of what is posted here

https://github.com/inktel/Vicidial/blob ... ICTIVE.txt

If this is exactly how the dial level is calculated then thats fine, i just wanted to make sure I got the most recent algorithm/calculations that the predictive dialing uses to calculate a dial level, the github post above is pretty old.

I can't say I've ever seen ADAPT_AVG Dial Method specifically bring down a VICIdial system by itself. It's usually just a symptom of other issues, not the actual cause.

We've had clients regularly dialing 2,000,000+ outbound calls a day on a single properly-configured VICIdial cluster(using the ADAPT_AVG Dial Method) without issues before, so "high-volume" isn't usually a problem, if the system is sized correctly.

What exactly do you mean when you say you're "dialing at a really high volume"?

As for the PREDICTIVE.txt document you mentioned, it's in the VICIdial source code and hasn't been removed, although it hasn't been updated in 15 years. Have you looked at the source code for the current adapt script yet?

mflorell wrote:I can't say I've ever seen ADAPT_AVG Dial Method specifically bring down a VICIdial system by itself. It's usually just a symptom of other issues, not the actual cause.

We've had clients regularly dialing 2,000,000+ outbound calls a day on a single properly-configured VICIdial cluster(using the ADAPT_AVG Dial Method) without issues before, so "high-volume" isn't usually a problem, if the system is sized correctly.

What exactly do you mean when you say you're "dialing at a really high volume"?

As for the PREDICTIVE.txt document you mentioned, it's in the VICIdial source code and hasn't been removed, although it hasn't been updated in 15 years. Have you looked at the source code for the current adapt script yet?

I just found it, the perl script correct? I'll dive into it and put it together thank you. Maybe it's something else. I was mostly referring to the number of outbound calls, but if there are other companies running that high than I can imagine it must be because of the other things we run in conjuction with vici such as the api_calls that we make to automatically insert leads into lists via hubspot and other things that were developed before I got here. I've only been working with the system for about 3 months and I was trying to get ahead of the game by improving and alleviating load on vicidial since during my time here, we've had vici break down on us plenty so I'm sorry if I sound unexperienced.