In short, the aim of this project is to develop a commercial, hosted PBX platform based on open source components. The users of this platform will be telcos, distributors and end user companies.
The project is to be developed as an open source project by volunteers, who initially participate “con amore” anticipating the opportunity of the future consultancy revenue.
Some people might say that there are lots of hosted PBX products! My answer is no: There are no Class 5 products except for the “incumbent” players such as Cisco, Avaya, Nortel and a few newcomers such as BroadSoft. Furthermore, there is no suite including all the features required by a new operator or a big corporate customer: Customer and user management, site management, provisioning, APIs, multi language capabilities, legal interception, change tracking, etc.
Options not described yet: Aim at a world-wide network like Skype. Include a feature full billing system.
A more detailed description
1.1 Market target
1.1.1 End user perspective
Small to medium size enterprises: 10 – 5000 end users. The geography of the customer is of minor importance: it may be a highly concentrated customer, as well as a customer with many small offices.
1.1.2 Distributor and Telco perspective
The distributor is the traditional distributor of PBXs including telecommunication operators. The system may exist as a stand-alone system without a distributor, or as a stand-alone system managed by the distributor.
1.2 Product description
The core of the product as seen from the end user’s perspective, is a PBX with the mainstream features of PBXs. In comparison, the feature set will be limited according to the 80/20 rule.
But in addition to the standard PBX feature set the system will have:
- Intuitive graphical web interface for managing the system
- Multi-locale (language, culture) design of all interfaces
- Interface to several commercial PC-based switch boards
- Plug-ins for the main Office automation programs
- Interface to LDAP allowing for deployment of new extensions from active directory
- Basic mobile telephone integration for SIP-enabled mobiles
- Open APIs for third party management programs (possibly including CSTA)
- Open APIs for third party call control programs (possibly including CSTA)
The web interface and the plug-ins will match or exceed the standard of competing products from e.g. Avaya or BroadSoft.
The topology of the voice platform is designed to cope with the challenges caused by a typical Internet with packet loss, latency and throughput problems. It will be distributed on many small self-contained units controlled by a central management system. The central management system is either owned by the end user or the distributor/operator.
A number of components useful for the distributor or network operator of the product include:
- Management system for deploying new customers including the management of change requests
- Management system for keeping track of the components in the back-bone network as well as the components located at the customer premises
- Routing backbone network for interconnection to the PSTN and other networks
- Central CDR for the recording of traffic and the use of features
- Multi-locale (language, culture) design of all interfaces
- Legal tapping system
- Central logging of changes made by the distributor/operator or the customer
The part of the system that is designed for the distributor or network operator is not mandatory, thus the system may be sold as a stand-alone system located at the customer premises. On the other hand, the components described above make it possible for the distributor to provide a better degree of support to the end customers and in addition be able to benefit from the role as traffic provider.
1.3 Open source
The platform will be based on several open source components: OPENSER, FreeSWITCH, MySQL, PHP, Cacti, etc.
What forms a platform from the components is “glue” consisting of:
- State of the art documentation of all components
- An architecture and implementation of a distributed system
- A strategy for a high-availability (HA) architecture
- Common management system (possibly developed in PHP), multi language
- Programming of OPENSER, FreeSWITCH and possibly Asterisk
- A license management system
- APIs for call control, provisioning and network management
- End user and distributor WEB-interfaces
- Documentation on all levels
- Marketing material
- Training material
- Best practices manuals.
A main challenge is to develop a license management system which protects the owner of the product at the same time allowing for flexible price structures.
The system may also include a minimum-feature billing system allowing for a quick start, but the goal is not to develop a fully fledged billing system.
1.4 What will the sources of revenue be?
The following sources of revenue are envisaged for the project:
- Training
- Maintenance
- Tailor made solutions
- Turn key project management.
1.5 Best practices development
The mentioned components have already shown their capabilities and limitations. Nevertheless, there is experience in the market place in how to use these components in a way, which secures the stability of Class 5 telco products.
Even within the limited market of Denmark, many VoIP systems exist based on the mentioned components. There is no reason not to assemble the specific know-how gained in the companies and to base the core architecture on the collective knowledge gained by the market players.
It could be tempting to expand the scope to other countries thereby expanding the collective knowledge, but the gain is likely to faint in comparison with conflicts of interest and conflicts of intellectual property rights.
1.6 Time and cost to market
A tentative duration for the time to develop these systems is 24 months:
- 6 months to gather the open source team, define common tools (wiki, documentation standards, testing, bug reporting) and define the product and its architecture
- 9 months for the core development
- 12 months for extensive testing, development of APIs and plug-ins, documentation and the development of training materials, including time to gain the first market experience.
A core team of 2-4 people with expert knowledge and experience is likely to be required for the first 12 months. The team must cover areas such as: Linux expertise, database, PBX, SIP, CDR, management and web design.
The total cost in terms of manpower is likely to be around 15 man-years.
1.7 More information
… to be published in other pages on this weblog.