In 1993, Arash Afrakhteh immigrated to the U.S. from Iran to work at Microsoft. He initially felt like “a kid in a candy store,” as a result of the stimulating intellectual environment, bright people, and excellent software tools at the software giant. Arash later worked at Apple, Broadbase Software and at Vertical Networks where he was the principal architect of their award-winning network management software.
In 2001, Arash became an entrepreneur when he co-founded Cariden Technologies, Inc. without taking any money from VC’s or angel investors! A team of about six people were able to get the self-funded company profitable in 2003. The exit came on December 17, 2012, when Cariden was acquired by Cisco for $141 million in cash and retention-based incentives.
In the press release, Cisco stated, “Cariden’s products and technology will advance Cisco’s nLight technology for IP and optical convergence and will allow service providers to improve both the programmability of their networks and the utilization of existing network assets across the IP and optical transport layers.”
Later, Cisco said its acquisition of Cariden “will allow service providers to enhance the visibility, programmability and efficiency of their converged networks, while improving service velocity.”
Arash told his amazing story of entrepreneurship and start-up success at a well attended TiE-Silicon Valley event on August 6, 2013 in Santa Clara, CA.
Within a few short years of its 2001 creation, Cariden quickly became a respected developer of IP/MPLS capacity planning, traffic engineering and management software. It’s now known as the “MATE product family.” MATE software provides IP/MPLS traffic management solutions, which include maintenance planning, congestion mitigation, troubleshooting, and network look-ahead; what-if analysis, design verification, and strategic traffic engineering; and new-customer impact and capacity planning capabilities.
Arash said that 85% of global tier-1 ISPs currently use Cariden’s software to ensure network visibility, reliability, efficiency and performance while improving service velocity. Other Cariden customers include PTTs, MSOs, mobile operators, and many other operators of large networks in both the public and private sectors.
Last year, Cariden was able to leverage their network software expertise to create Infrastructure SDN for the emerging software defined networking market.
A Perspective on How to Succeed as a Network Software Start-Up:
Arash’s story was incredibly impressive and inspiring. It detailed how a small team of founders were able to understand the opportunity and challenges of developing and successfully selling “back-end” network management and traffic analysis software to different types of network service providers (SP’s). Arash currently serves as Chief Technology officer of Cariden Technologies, Inc and was previously its Director of (Software) Engineering.
Three major SP problems were identified by Arash:
- Control network infrastructure costs, but continue to manage capacity in order to support increasing bandwidth from more users.
- Urgent need to bring a “business approach” to the network (rather than “if we built it they will come and generate increased ARPU).
- Dealing with network/equipment failures and coping with “a changing dynamic matrix” of network infrastructure hardware and software.
In light of the above, Cariden’s mission was to provide visibility, analysis and control across the SP’s entire network infrastructure. This was a very real challenge for the young company as SP’s will generally NOT allow network software companies (like Cariden) to “touch their network.” So Cariden worked around that problem by obtaining “off-line access” to SP data. In 2001, Cariden built a MATLAB prototype and then created real software that was impressive enough to attract Deutsche Telekom and SBC as early customers.
Mr. Afrakhteh gave four examples of similar network start-ups (no names were mentioned) that either failed or were acquired at much lower valuations. While each of those companies had their own unique set of problems, the one that stood out was “falling in love with your PhD thesis.” Evidently, that was the case with one start-up which was striving for an ultra-optimized software solution based on one of the founder’s academic research.
In 2011, the company began to focus on the embryonic SDN business opportunity, as championed by Jeff Bazar, Vice President of Strategic Product and Corporate Development. The company’s “Infrastructure SDN Platform” exposes network resources for continual optimization and predictable handling of diverse traffic demands. Cariden’s vision of abstracting and controlling physical resources through software — realized in the creation of a Network Services OS (NS-OS) — forms the core of its Infrastructure SDN solution.
It was quite clear from Arash’s talk that Cisco did NOT acquire his company to get into the SDN business! Instead, Cisco said, “Cariden’s widely deployed and industry leading capacity planning and management tools for IP/MPLS networks will be integrated into Cisco’s Service Provider Networking Group (SPNG) to drive multilayer modeling and optimization of optical transport and IP/MPLS networks.” It remains to be seen how and if Cisco will use Cariden’s software for the SDN business, which many feel threatens Cisco’s dominance of the switch/router market.
Arash Afrakhteh’s note (in an email):
“The MATLAB prototype was our original prototype in 2001 for VCs and not related to SDN prototype. Same for DT and SBC, they were early customers and we used real software and not a prototype to work with them back in early 2003. Our SDN prototype was a distributed scalable SDN solution and it was in its own right compelling in a market that lacks real world solutions.
Also, Cisco may use our software for SDN, I can’t comment on that, but I did not intend to communicate that SDN was not on Cisco’s mind when they acquired us.”
During the Q & A, Arash was joined by Alan Gous, Vice President of Product Management at Cariden.
The first question, from this author, was how Cariden could break the “not invented here syndrome of large telcos which were determined to provide their own software for capacity planning, traffic management and analysis. Arash and Alan admitted that was a very difficult problem. However, with a few carrier connections (Global Crossing plus several unidentified telcos) and a convincing network software solution they were able to get their foot in the door. When it was clear that the Cariden software was superior to any “home-grown” carrier software generated tools, it became an easier sale.