Realtime Interview with David Mandelstam at Sangoma Technologies
In my ongoing efforts at talking to interesting people in the VoIP sector to telecommunications, I've encountered some folks with incredibly interesting stories to tell. And some with keen insights into how the technology has evolved from the old legacy of TDM-based telephony or how it is evolving into the next generation beyond today.
The other day I had the opportunity to chat with David Mandelstam in Toronto. David is the President and CEO of Sangoma Technologies.
David's Background
David and his research and development team focuses on Sangoma’s family of AFT (Advanced Flexible Telecommunication) T1/E1/J1 voice/data cards that are engineered for today’s demanding soft PBX, IVR and VoIP applications, such as Asterisk, Yate, and OPAL.
He recently spoke at the VON Spring Conference & Expo, San Jose, CA about Free and Open Source VoIP: World Communication through World Cooperation
A Little About Sangoma Technologies
Founded in 1984, Sangoma Technologies Corporation is a leading provider of connectivity hardware and software products for VoIP, TDM voice, WANs and Internet infrastructure. The company develops and manufactures voice and data communication products including the industry-leading series of Advanced Flexible Telecommunications (AFT) PCI cards. Sangoma Technologies Corporation is publicly traded on the TSX Venture Exchange (TSXV: STC - News).
Earlier this year, Sangoma was named to the VoIP Magazine 20: Companies to Watch in 2006. The VoIP Magazine 20 features the companies that are best positioned in the coming year to spur IP communications adoption and innovation while, at the same time, delivering on the immediate needs of the marketplace.
The Sangoma Voice solution is described like this on their web site:
Here's one visual of how Sangoma voice can fit in a VoIP architecture -
All Sangoma’s voice related AFT products, from the A200 FXO/FXS analog system through the A101 (single T1/E1), A102 (Dual T1/E1), and A104 (Quad (T1/E1) :
- Use the same base PCI interface card
- Use the same professionally engineered firmware on the same family of Field Programmable Gate Arrays.
- Comprise an integrated family of products that all work identically in all motherboards under all conditions of shared and unshared interrupts all the time.
- Conform to 2U specifications and all are delivered with both normal size and 2U brackets. There are no mechanical interference surprises when you use Sangoma products.
- Are delivered complete with high quality, tested cables.
Sangoma’s PCI architecture (of course with autosense 3.3/5v support) has superior performance and compatibility simply because the family approach means that we only ever have to solve an interface problem once. If you don’t enjoy experimenting with different motherboards, then this family is for you.
Ken: Sangoma has a very different fit into the VoIP technologies sector than many of the companies I talk with. Many of them are directly engaged, whereas Sangoma actually develops hardware and partners with some of those players. Can you tell us a bit about Sangoma and help set the stage with our readers? I know you do some work with Asterisk, but I sense much more under the surface.
David: I'd like to stress that we have a good overview of open source. We're somewhat unique in that we're not tied to any one signle open source project. We work with many of them. For us, the focus isn't so much Asterisk as it is open source VoIP solutions. Asterisk is a very visible open source solution, perhaps the most visible. They've had great success . They're the olders and have reached a self-sustaining level of market penetration. We also work closely with Freeswitch,Yate and OPAL. (Note from Ken - Yate is Yet Another Telephony Engine. OPAL is Online Programming for All Libraries. Both are popular open source VoIP alternatives.) Open source solutions represent, for us, the entrepreneurial, rapidly growing center of innovation. I believe Yate has also reached that self-sustaining level that ensures ongoing growth and success.
The ease of integrating and customizing open source technologies is unprecedented for developers working on solutions. We could always develop custom solutions, but in the past you hired a specialized programmer who lived on Pepsi and Twinkies doing soem programming magic in a balck box of sorts. Nobody understood the custom code and it took very specialized skills to develop. With todays open source solutions and the talent in the field, customizing open source code for Caller ID, IVR and many other features are being done routinely.
The PBX or key system were always viewed as fully featured devices, but that's a fallacy. There are things they just can't do. Call coverage is a good example. Users are always locked into pre-defined call coverage. For example, if I'm not at my desk the call goes to voice mail. If you're lucky you can hit 0 and escape to a receptionist, but it's a fixed finite path. Open source coupled with an IP PBX solution can easily allow options. Callers can get a menu and pic whether they want to try my cell phone, or go to my boss. Or I can set a coverage path that tries the receptionist then prompts for the caller to decide on voice mail or a live person. If needed, I can ring every phone in the office simultaneously to ensure the caller gets a live response. Traditional systems just can't offer that kind of flexibility.
One are where we're quite different is we sell to anybody. Individuals, small business, large enterprise. Our customer base includes resellers, integrators, developers, VoIP service providers and many of our customer still do a lot of traditional telephony. We're different because we are traded publicly. Many companies involved in VoIP are conglomerates. if you invest in Cisco, Nortel or Avaya, you're investing in the corporate machine. You can't control how your investment is used at all because they are so large and diversified. Sangoma gets our investors closer to the technology they want to invest in.
We find some of our happiest customers are people who are building systems in volume. We're well know for the quality of our hardware. Our reputation is built purely on quality and that's what brings our customers in. They want bang for their buck. These people building systems in volume are buying and selling peace of mind. We help deliver that and Asterisk has lowered the barrier to entry for providers who want to deliver a customized niche solution in volume.Ken: I'm curious what you see as the acceptance of open source in business. I see small companies everywhere embracing it, but what about large enterprises?
David: Asterisk is mostly used by relatively small companies. There are many of these companies. Too many to count. Asterisk represents and affordable system that can meet their telephony needs. Whether it's a call center or a small PBX for an SMB company, Asterisk offers a good solution. We're also seeing Asterisk used widely for providing what I'll call voice chat lines. Vocie service people can call into and chat on the phone with other callers, not unlike Internet chat rooms. What we see is product in volume, based on Asterisk now being produced.
Open source VoIP in business has a pretty visible comparison directly to Linux. It wasn't that long ago that Linux wasn't anywhere in the business sector. It just wasn't used. It was considered too risky. But then we saw Linux become the platform solution for market products. Linksys routers. Tivo. Firewall appliances. When Linux become the OS for delivered platforns, it surged in business and is now prevalent as a server OS in the business world. It's everywhere today. Open source VoIP appears to be following a similar model. It's being accepted, but just now is becoming the platform for products in volume. Asterisk is becoming a framework for delivery if commercial IP PBXs. Success with those products will drive widespread acceptance and adoption.
We certainly see large companies embracing open source technologies as a part of their telephony solution. Call centers based on Asterisk are becoming a huge market. In a large enterprise there may be a business unit that needs a call center for a specific purpose. Asterisk is a good fit that we see many large companies using more and more. Again, it's the quick deployment, flexibility and customizable nature of open source they like. We see Asterisk used for custom IVR solutions.
We're beginning to see 1000 line call centers based on Asterisk. Some bigger systems are built on Yate. Big conference bridge systems.
David: Absolutely. Screen pops are trivial and done all the time. We're seeing some pretty innovative integrations. We're seeing some pretty big CRM systems built on Yate to integrate voice and data together.
Ken: How healthy do you think the market is for people trying to build on open source solutions?
David: Some people are making a lot of money people who are building systems for mass rollout based on open source are making a lot of money. Integrators who install customized systems in vertical markets are doing pretty well. People who are working at installing individual systems are struggling harder.
Ken: So what do you see as a big system? Is there an upper limit?
David: There are a couple of ways to look at that. A 1,000 line call center is a pretty fair sized implementation even in the traditional PBX market. The fact that open source systems are doing that sort of thing says they're doing well.
I know of a Yate system that currently processes 480 simultaneous calls on a single computer. That's pretty good for a PC and they're looking to double that capacity shortly. The thing about VoIP and open source is how it lowers the barrier to entry. Traditional PBXs and key systems have always been very expensive. PCs are pretty cheap these days.
A PC is a fast enough processor to do most of the call handling work these days. There are a couple of areas where PCs aren't the best. Compression and decompression, or the transcoding that's need to go from a G.729 codec to a G.711 code eats up a lot of CPU power. Echo cancellation is also a very processor intensive task. And while many of the computers used today might be a 32 bit or 64 bit,processer, remember that voice coding is all built to an 8-bit structure. 8-bit maniupalation on a 64-bit processor isn't making the most efficient use of the CPU. But for most VoIP related tasks, a PC CPU can do a very good job of carrying the workload.
Some operations could be moved off the PC to another processor. That allows the pC to operate more efficiently and effectively. We think we help provide the optimum mix od off-PC handling for TDM voice integrated with the VoIP environment. And there's even an SS& application, so full integration with the PSTN is being accomplished right now.Ken: David, I have to admit that you've given me a lot to think about with regard to open source in VoIP. I know I've got some biases for enterprise business based on my own telephony background, but you've really given me some new ways to think about how open source solutions from Asterisk, Yate, OPAL or any of the open source projects, might really be used inside an enterprise to augment the existing telecommunications services. Do you have anything else you'd like to add?
And again, we're publicly traded, so if people are interested in investing close the technology enabling VoIP, we welcome their interest.
Wrap up from Ken
I have to say that talking with David was, in many ways, an eye opener for me. I have a pretty healthy respect for the Asterisk project. I have a system up in my lab in my home office. But I saw it as a small business replacement for a key system really. I truly didn't see it as bringing much more to the game.
Understanding the ease of cranking up a custom call center inside an enterprise, whether it's for a help desk, customer service, taking orders, or any other purpose, brings an increased respect for the depth of workmanship. And I have to agree with David. packaged systems, delivered to market in repeatable volumes with quality will move open source from a niche product to a framework. Once that chasm is crossed, open source VoIP will make significant, aggressive inroads into the enterprise business space.
As with every interview I do, any errors in this writeup are mine, and not due to any oversight or ommission on David's part.
I'd also like to thank Sheldon Rose at Sacke and Associates for helping get David and I together.I don't believe I have another interview scheduled at the moment, so I'm not sure who might be next. I know I'm working with the folks at VirtualPBX to see if we can find a good fit for what they do on here. I'll be working on that ahead. And a brainstorm I mentioned a time or two recently is quickly forming into a group you'll hear from soon that I think of as The VoIP ThinkTank.
If you find these interviews useful or helpful, or there's someone in particular (either an individual or a company) that you'd like to read or hear an interview with, please drop me a note.
For archival and reference purposes, a PDF copy of this interview write-up will also be available in the Realtime VoIP Community Reading Room.
Technorati Tags
VoIP
IP Telephony
Voice Over IP
Sangoma Technologies
David Mandelstam
VoIP Community
Realtimepublishers
Realtime
Realtime VoIP Community
Email This!
Digg it!
Sphere it!
Del.icio.us
Reddit!
Newsvine