Contact centers are becoming increasingly distributed and are beginning to leverage hosted VoIP services to drive efficiency, automation, and distribution into their communications infrastructure. However, due to quality management, regulatory and legal considerations, most contact centers are finding that they cannot completely outsource their communications to a hosted VoIP provider because they cannot maintain features such as call recording. Customer premise based recorders require complicated networking changes for the customer and are not able to handle the increasingly distributed nature of the contact center.



VoIP communications relies on the transmission of “packets” that contain content of the call. One VoIP call can contain numerous packets and in the course of transmission from one origination to terminating destination, the packets are generally received in random order. However, each packet contains information in a header, payload, and trailer section relating to the source & destination, the actual content of the call (data), and information identifying the order in which the packets should appear so that the transmission makes sense.

The information contained in the packets that identify the order in which they are to appear, or the “instructions”, are also known as protocols. In very simple forms, protocols can appear individually. However, more commonly, protocols appear as stacks, multiple layers of protocols that work together to form the resulting outcome.

The most widely used protocols for VoIP communications today are H.323 and SIP. Although, all voice and video communication is generally performed over a separate protocol known as RTP (Real-Time Transport Protocol). RTP defines the standardized packet format for delivering the aforementioned type of communications over the internet.

In the past, when customers replaced their PBXs every 3-7 years or so, the expense of new physical equipment was a given. But now, with the entrance of VoIP onto the scene, it is possible to deliver all the functionality of hardware-based, key systems for much cheaper—using software. In order to compete in this new marketplace, service providers need to be able to offer hosted solutions to their customers—solutions that require little or no up-front investment, but instead are paid for through an ongoing contract. It is also important, to compete in the VoIP marketplace, to continue to provide customers with services that were once important in traditional telephony in the VoIP world as well.

Examples of such service are call accounting and call recording. Contact centers want applications that can record calls for quality management, regulatory and legal compliance issues. Better yet, contact centers want hosted call accounting and call recording products that meet their needs and cost them little or no upfront capital investment.

In a hosted, call recording scenario, VoIP extensions that require recording are registered on the Service Provider's switch. A VoIP call is placed by the contact center agent and a SIP invite is sent to the switch. The switch recognizes that that traffic for this extension is to be recorded and redirects the SIP invite to the call recording server. The SIP invite is then sent back to the switch and either forwarded on to the terminating VoIP address or to the PSTN (Public Switched Telephone Network) for termination to an off-network user. Once the SIP invitation is accepted, packets of data containing the call content begin to be exchanged.

This call content is exchanged in the form of RTP packets. These RTP packets are passed from the call originator, in this case the agent, to the call recording server where they are recorded and passed on to the termination point. Call content from the termination point similarly passes from the termination point to the call recording server to the originator. The call recording server transcodes the call content into an MP3 (Mpeg 2.5 layer III). These MP3 files are then sent to a file storage/web server for access via a browser-based application.

The browser-based application is able to display traffic for calls currently in progress as well as calls that are stored on the file server. A feature rich browser-based application will offer the ability to display calls based upon user permissions as well as offer features such as call descriptions and filtering. The browser application should also offer the tenant the ability to decide parameters and thresholds for call recording such as by hour, day, number, or a percentage of call traffic.

VoIP call recording options exist in several forms. The purpose of this paper is to address those platforms offered by the hosted, VoIP Service Provider. Along with packet routing call recording technologies outlined earlier, technologies exist which involve the practice of capturing or “sniffing” data packets as they travel through the Service Provider's network. In this technology, all data packets carrying signals or content are reviewed by the call recording application. As each packet is touched, the call recording application decides if the packet is to be recorded. This technology can be challenging for the hosted Service Provider due to the inherent scalability problems of touching every pack.

Additionally, situations exist for aggregators of VoIP services who only wish to offer call recording services to Service Providers who have purchased the service.

Challenges also arise in recording extension to extension calls placed within the tenant as these calls do not typically leave the customers network and are not routed through the Service Provider's switch.

Call recording is an integral part of the contact center environment as well as required for many regulatory and legal purposes. VoIP service providers offering value added services including call recording expand their reach into lucrative vertical markets and retain customers.