How Difficult is It to Host Your Own SIP Based Dialer?

Right now I'm using vicibox redux on a celeron piece of crap box I got for 150 bucks trunked thru vitelity at 1.44 a minute for 21 lines max (I'm not positive my box can run all 21, not tested). It handles 8 just fine though.

I already have the fields and process worked through logic hooks and addons and php scripts to push the leads and drip marketing down the tube.

I'm in like a usable alpha state now. I have 1 script left to fix that does some mySQL selects and updates to change statuses using the process manager addon to call external scripts (its not the best solution, but this is just a proof of concept box).

After this is done I'm gonna beta this out a couple months and then redo the hooks using a custom module instead of process manager.

The dialing is the fun part, I'm actually dynamically scaling the lines up and down. Completely flies in the face of the normal call center philosophy, but I'm doing this from a individual agent working a book perspective, not a call center model.

Do you have to license goautodial? Vicibox is a free iso done in GPL v3 public domain license.

I haven't used goautodial, but I think best I understand they're both working on same asterisk variant, seems like goautodial runs with ubuntu instead of opensuse tho, not sure if there is anything else different.

Vicibox comes with a vtiger install that it is already integrated with, but I didn't even try looking at it, basically because I wanted to start with a newer code base and I do not care if it says my dialer is vicidial and my CRM is sugarCRM on the pages, I actually like having the credit to the developers that came before me on there.

Like I said, the reason I'm doing this is just because I couldn't find anything out there that did what I was looking for so I just started making it in my spare time expecting it to take a LOT longer. Now I have a working alternative that is going to save me 250 bucks a month over what I was paying for CRM/dialer, and it has a better dialer.

Goautodial is cc. I was referring to insurance licensing.
 
Fascinating thread. I never knew you could run your own "call center."

For those of you who want to run Linux, it's not a big deal to partition your PC hard disk to run one of the flavors of Linux (there are like 300 of them... see DistroWatch.com: Put the fun back into computing. Use Linux, BSD. for a list.) I'd suggest Ubuntu as it is the easiest to install and the most popular... although I like Arch better... but not for newbies!)

Or maybe you want to just run Linux in a virtual environment by using something like VirtualBox which is free... and quite good from what I understand. (I use Parallels which is overpriced IMO compared to VB (free !) but there are tech reasons why I don't change... too long to explain here.)

Ksig, if you ever get your turnkey package complete, I'll bet lots of agents will buy it. Talk about something really worth $300... or more!! (IMO... YMMV.)

I don't know the first thing about dialing or telephony. What I'd like is just some software (for Mac or PC) that will let me load a bunch of phone numbers in it and let me connect it to my telephone (but how?) and dial on my line until someone answers. I just want "simple" and so I don't have to punch all the buttons.

The CRM addition is nice... Sugar CRM is a good platform because it is "open" and you can play around in the code as well as find lots of developers who can help. As for being "rock solid" well... not always. An update comes out about once every six weeks or so. Best to wait a week or two to find out what the bugs are from the early adopters. Usually the bugs are in the more esoteric areas of the system... things I never use.

Keep us posted on this "journey" of yours. You might just come up with something that we will all buy and you can displace a certain arch-conservative Republican vendor here as the wealthiest subscriber in the forum!! :laugh: :yes:

Make it happen!

Thanks for your updates.

Al
 
I finished up the final touches on my dialer today to where I'd call it usable.

It now imports my leads into sugarcrm via post/get from all sources, loads them into the dialer, starts drip marketing, reloads the leads in a pattern so they aren't overcalled yet are attempted a good number of times, and dynamically scales the phone lines up and down intelligently for single person.

And Al, tbh I'm considering giving away the turnkey package in ISO format for free with no support, then providing hosting for people who would rather have it hosted, then charging about 30$ an hour for any custom work people want (scripting and integrations).

For example, I'm going to use sugar to iframe in quote engines using the code to pass out a get variable of the currently open lead ID, then parse the mysql of that lead id to get my variables to open a quote page of the lead. First test is going to be with compulife, because his CGI is really easy to work with.

The only thing I don't want to give away is the sales process stuff that is inside the algorithm for the dynamic phone line scaling and call attempt timing.

All I'm trying to do is patch in a middleware between all the stuff I use so I can use it in 1 place without copying any data to make myself a lot more efficient.
 
Do you run into any issues with a dynamic IP vs static IP?

The folks I use for my SIP dialer (which runs like swiss watch) do that exact same thing. They give away the software for free, but charge MUCH more than $30/hr for support.

www.HostedPredictiveDialer.com
 
Last edited:
You can buy a dynamic ip dns reg for like 15 bucks.

Mine hasn't changed in forever, and if it did I can either buy a dynamic one or cronjob a script to find the current ip and post it back to my crm once a minute into a configuration file.

Either option would work. I'm sitting next to the dialer so it would have no effect on my calls but could kick an external agent maybe once every couple weeks for as long as 1 minute. Probably wouldn't even notice.

Once I get to something more serious I have a colo center that is able to support with static ips.
 
I was using a colo site, but I moved and the local cable provider was able to give me a static IP address ("business class") for a lot less than a a colo site. I also have (2) 4U servers and (1) 1U server, so that was driving the price on the colo site too.
 
Lucky for me sugar has very limited overhead. 1 dedicated server could probably handle 20-40 average users or more. At the very most it pulses up like 400 megs of memory usage on lead creation in my system, and I have it performing like 20 actions on a new lead at once that include tons of sql writes and posts, and they're being done with a 3rd party software.

I could probably write something similar that would be more efficient if I was given enough time.
- - - - - - - - - - - - - - - - - -
I also would like to point out, I started this project the day I posted this thread. For someone who had never done this before, with just help from a developer that cost me around 300 bucks, I'm up and running. Total cost less than 1000$ counting equipment and all.

It's funny what you can do with off the shelf open source software with a bit of modification and a book on php/mysql + the online reference manuals.

For the record I did have 2 years of college computer science classes and 15 years of video game experience writing my own scripts and stuff, so I'm pretty sure you aren't going to see just every guy out there running out and building something from scratch without a couple snags here and there, but it's a hell of a hobby.

Some people build ships in a bottle and go fishing... I build scripts. I think I'm going to see just how much of insurance sales I can automate as a process with no human interaction necessary.
- - - - - - - - - - - - - - - - - -
And I wish I had listened to Al earlier about Sugar.

I made the dumbassed decision that an open source CRM wouldn't be as good as what the paid providers were supplying. I was wrong. I firmly believe at least 2 of the big CRM companies are using a reverse engineered or private labeled variation of sugar or vtiger now, after having experience with their CRM systems and seeing what sugar does. The similarities in how they function are very obvious.
 
Last edited:
And I wish I had listened to Al earlier about Sugar.

I made the dumbassed decision that an open source CRM wouldn't be as good as what the paid providers were supplying. I was wrong. I firmly believe at least 2 of the big CRM companies are using a reverse engineered or private labeled variation of sugar or vtiger now, after having experience with their CRM systems and seeing what sugar does. The similarities in how they function are very obvious.

I've been advocating SugarCRM here for the past five years. People say it is too hard to install and that a free system can't be as good as a paid-for one. I don't think that is true for everyone, but I grant that for some who are totally computer illiterate, they might be best advised to buy Salesforce or ACT.

True, Sugar is not expressly for insurance but it is not hard to "make it work" for the insurance industry.

I get a lot of heat for suggesting free or low-cost alternatives to CRM systems or quote engines or virtualization software. The prevailing theory is that if one uses "free stuff" they must not be very successful in the insurance biz. (How many times have you-know-who made that claim about me!!!)

I simply believe in keeping overhead low and giving low-cost products and services a try before you drink the kook-aide of the vendors here and spring for big dollars for their products and services.

Sure, some products and services are going to be efficacious for you... but at the end of the day none of them are going to be the magic bullet. You are still going to have to get out there and prospect and sell.

The vendors of various associations and quote services and training systems and FMO selling systems and lead companies all try to convince you that you have to pay big dollars to be successful when the truth of the matter is that you can be successful without any of them... but by plain old ACTIVITY.

The guys who hawk quote services and leads and associations and FMOs all say they have "what you need" to make it. Well they remind me of the guy who knows 40 ways to make love... but doesn't know any women!!

My suggestion is to save your money on what the vendors here are trying to sell, and instead get out there on the phone or in person and meet prospects. THAT is what is going to make you successful, not the couple of lead you might get from a quote service or some video from an association. True, they won't hurt, but there is only so much time and so few dollars.

SugraCRM is free. Virtual Box is free. Linux is free. Cold calling and cold walking is free. Try the low-overhead route before you spend your hard-earned money.

That's NOT the message the vendors want you to hear... but at some level I'm sure most of you will agree with me.

YMMV.

Al
 
Vicibox is free, sugar is free asterisk is free. Anyone can do exactly what I'm doing.

Sugar is sorta tricky to install on a remote server because of the chown and chmod issues. My hosting company gave up and gave me root access to let me fix it myself.

There isn't anything that prevents you from renting a host to someone with modified sugar ce on it.

I just love that it's open source. You can literally make it do anything because of that.

I'm about to write a php script to check all the leads where status is converted and email me 7 days before their birthday to mail a card.

Gonna make it load them into the dialed on their birthday too.

I asked leads360 for that feature 4 months ago and they didn't respond. Asked them to integrate google calendar too, no response.

I gave them most of my ideas and they just ignored them. That's part of why this ended up happening. I couldn't get what I was looking for exactly so I got a book on php instead.

Now I have what I wanted. It's going to be slick once I build compulife and quotit into the crm as a click to quote. Will save minutes of clicking per quote generated.

I suspect I'll be able to work leads twice as fast with this than I could before. More contacts = more sales.
 
If you have a one step ISO ready to go I'd love to play with it a bit. Does your flavor incorporate a timing device or would more code/work need to be done for that?
 
Back
Top