Eisodus Networks Pvt. Ltd.: A Tale of Unsuccessful Commercialization Attempt by a Start-up

Founding of Eisodus:

Eisodus was founded in October 2002 by Professor Abhay Karandikar, Sunil Mehta and Abhijit Gadgil. Initially, Prof. Karandikar officially joined as the chief technical advisor to the company. He could not become a full time employee of Eisodus as he was a faculty member at Indian Institute of Technology (IIT) Bombay. He finally joined full time from January 2005 after taking a year long sabbatical leave from IIT Bombay.

Background of the Founders: Prof. Karandikar obtained his BE in 1986 and MTech and PhD from IIT Kanpur in 1988 and 1995 respectively. During 1988-89, he worked for the Indian Space Research Organization (ISRO), Ahmedabad, an Indian government funded laboratory engaged in research in the field of space applications. He was a member of technical staff and team coordinator in High Performance Computing Group (HPCG) of Center for Development of Advanced Computing (C-DAC), Pune during 1994-97. Since 1997, he has been working in the department of Electrical Engineering of IIT Bombay as a faculty member. Prof. Karandikar had extensive consulting experience in the area of networking and communications technology and was actively engaged in research related to quality of service (QoS) guarantees in Internet and architectures for next generation networks (NGN).

Sunil Mehta obtained his BTech in Aeronautical Engineering in 1982 from IIT Bombay. In 1983 he joined MBA programme offered by Carlson School of Management at the University of Minnesota in USA and completed it in 1985. He joined his family business of transport and clearing/forwarding agency in 1986 and was also involved in the building of a chemical manufacturing business till 1989. In 1990 he started Telenet

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 1 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation.

Systems Private Ltd. which was involved in the development and manufacture of telecommunication equipment such as ISDN based modems and other customer premise equipments (CPEs). Telenet had its own PCB development plant based at Navi Mumbai near Mumbai.

Abhijit Gadgil completed his BE from Pune University in Electronics in1998 and obtained his MTech, working under the guidance of Prof. Karandikar in July 2002. He had experience in the development of protocol stacks and had successfully developed and released MPLS stack on Linux platform and a Linux based MPLS Emulator (LiME) which was freely downloadable. After passing out he joined Eisodus as a promoter with Prof. Karandikar and Sunil Mehta.

Indian Telecom Scenario:

National Telecom Policy (NTP 94) identified the importance of telecom development for the country and realized that the kind of investment required by the country in the telecom infrastructure could not be bourne out by Government alone. So it laid the foundation of a policy for the participation of private sector in the development of telecom sector in . National Telecom Policy (NTP 99) went a step further and laid down the direction for the Indian telecom development. A series of progressive steps such as the introduction of third and fourth telecom service provider in each circle led to considerable competition among the players. As a result the prices plummeted and high demand for telecom services was being witnessed during 2002-2003. Although the operators had laid down their equipment but the size of demand was such that they were constantly in the process of ramping up their infrastructure. The regulations also were playing a major role in the creation of new opportunities irrespective of the choice of the technologies. With near total opening up of access services to all for voice, video and data, new technologies were being developed to achieve convergence.

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 2 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation.

Broadband has been defined as an always on connection with download speed exceeding 256 kbps in India. India with its rising middle class and a large young and techno-savvy population was expected to bring in a very high demand for broadband connectivity. Recognizing the potential of ubiquitous broadband service in growth of GDP and enhancement in quality of life through societal applications including tele-education, tele- medicine, e-governance, entertainment as well as employment generation by way of high speed access to information and web-based communication, Government finalized a policy to accelerate the growth of broadband services and this was released as the Broadband Policy document in 2004. However, given the state of last mile connections across India which were mostly copper based many innovative products were required to provide the kind of service customers were looking for and this was seen as a huge opportunity by the new companies that were coming up. This growth was expected to drive a demand for network based products in the carrier space.

The Opportunity:

Broadband access industry, which was on a fast growth path, has been upgrading the technology constantly. The broadband access network comprises of first mile access and metro core network. The most popular first mile broadband technology at the time was Asynchronous Digital Subscriber Link (ADSL) with ATM (Asynchronous Transfer Mode) as the transport protocol for ADSL. The other alternate first mile technologies were Data over Cable Service Interface Specification (DOCSIS) networks and Passive Optical Network (PON). However all these first mile access technologies had some problems associated with them. For instance, ADSL had an issue of complexity of management and high overheads when integrated with Internet protocol (IP) based networks. The DOCSIS based cable network was associated with high latencies and low bandwidth per customer due to its shared nature. PON with ATM as the transport mechanism suffers from problems similar to that of ADSL.

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 3 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation.

For metro core network Synchronous Optical Networking (SONET) or Synchronous Digital Hierarchy (SDH) rings with ATM as transport are primarily used. Post 2000 there has been a significant shift in traffic patterns and data traffic forms a major part of the total traffic volume through both the last mile as well as metro core network. Time- Division Multiplexing (TDM) based technology is considered inefficient and consequently, a shift from TDM centric network to packet friendly IP based networks is slowly taking place across the spectrum of networks. The existing access networks had problems such as low bandwidth, inefficient usage of existing bandwidth, inflexible provisioning at granular levels, and high cost. So, there was a need to develop a product that could bridge these problems and further add value for the telecom companies and their customers.

Motivation:

During his MTech days Abhijit had regular interaction with Prof Karandikar and during these interactions they had came up with some interesting ideas related to carrier equipment especially with respect to India where last mile access was mostly copper based. Once Abhijit graduated from IIT Bombay, he and Prof. Karandikar were looking for ways to operationalize their ideas by setting up a company of their own. Meanwhile, Sunil Mehta came over to meet Prof. Karandikar for some consulting work as he wanted to introduce a new line of products for his company Telenet Systems Pvt. Ltd. which at that point of time was manufacturing ISDN based modems and some other telecom equipment. But Telenet Systems manufactured no high end products that could be used by the network providing companies such as the telecom service providers (Telcos) or the Internet service provider (ISPs). During the meeting Prof. Karandikar and Sunil Mehta realized that their views about the future of access products were closely aligned and they decided to join forces to pursue their ideas together.

Both Prof. Karandikar and Abhijit Gadgil had no prior entrepreneurial experience. For Prof. Karandikar who was to become a first generation entrepreneur with Eisodus the

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 4 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation. basic motivation was a chance to develop a high end technology product and a chance to create a successful company from IIT Bombay. Among the IITs, IIT Madras (the TeNeT group led by Prof. Ashok Jhunjhunwala) had successfully started and was operating several companies within the telecom segment. Prof. Karandikar was sure that such an initiative could be taken up at IIT Bombay as well.

Sunil Mehta on the other hand wanted technical expertise of Prof Karandikar and his students to be able to develop high end products and possibly take Telenet Systems to the next level of evolution. His idea was to merge the technological knowledge of Prof. Karandikar and the manufacturing experience of Telenet Systems Ltd. to come up with market dominating and technologically superior products.

Apart from the founders other people who joined Eisodus included MTech students of IIT Bombay who graduated in 2002 and were closely associated with Prof. Karandikar during their MTech days. These included Zainul Charbiwala and Praveen Kumar. Both were looking for an opportunity job that was both challenging and promoted learning. They talked to Prof. Karandikar and after being convinced about the nature of work joined Eisodus. Over a period of one year several other engineering graduates joined Eisodus namely, Balaji Kasal, Ranoo and Hrishikesh Gokhale.

Getting Started:

Incubation: In 1999, Kanwal Rekhi School of Information and Technology (KReSIT) within IIT Bombay had started its IT incubation programme to promote a culture of entrepreneurship among the faculty and students of IIT Bombay and to commercialize useful technologies which were constantly being churned out of IIT over the years. The above experiment had a profound effect on all the members of the IIT Bombay family and encouraged by the success of the experiment, IIT Bombay set up a full fledged science and technology business incubator. The incubation center underwent several

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 5 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation. structural changes during the period of 1999 to 2003. Finally, Society for Innovations and Entrepreneurship (SINE) formally came in to existence in 2004 to administer the activities of the incubator. The incubation center enabled its incubatees to get access to certain critical infrastructure such as office floor space, equipment like PC‟s for development work and also helped to promote networking among the budding entrepreneurs and possible VC‟s and other angel investors. In return for the above benefits, incubatees were to pay fees for the usage of the facilities at IIT and they may negotiate to give a small equity holding of the company to IIT Bombay incubation center.

It was decided to start Eisodus from the incubation center of IIT Bombay during the initial period till the company could raise enough funds and could became self sustaining (refer Exhibit-2). Eisodus became a part of the incubation process even before SINE came in to existence and was amongst the earliest companies to be incubated at IIT Bombay. The choice of Mumbai as the location for Eisodus was contingent on the fact that it was incubated at IIT Bombay and could access the relevant resources of IIT Bombay for development. Also the presence of Telenet Systems and its manufacturing center in Mumbai was thought as advantageous for Eisodus.

First Activities: The initial seed capital of INR 2 million was infused by Sunil Mehta and INR 0.3 million was contributed by Prof Karandikar to kick start the operations of the company. It was decided to register the company with the registrar of companies in , to arrange/borrow funds for meeting the working capital requirements of the company and engaging a CA firm to look in to the auditing needs and day to day accounting requirements of the company. A series of meetings were conducted over the period of December 2002 to May 2003 in order to decide on the direction that the development the company would take and prepare a comprehensive business plan for the same. The idea of developing a business plan was to solicit investment from the potential investors by making them aware of the technological and other business details such as expected market size and possible customers of the product being conceived. A very important

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 6 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation. design philosophy set right at the start was to use open source software and off the shelf components as much as possible for the development work. It was thought that this would help in keeping the cost down as well as increase the pace of development.

One common acquaintance of Prof Karandikar and Sunil Mehta was Dinesh Singh1. Dinesh was an alumnus of IIT Bombay and was involved with Hughes Tele.com which was acquired by Tata in December 2002 and renamed as Tata Tele-services Maharashtra Limited (TTML). Dinesh was extremely knowledgeable person with a long experience of the telecom industry abroad where he had worked for major telecom equipment companies. So, Prof Karandikar and Sunil Mehta thought that they could discuss their plans with him and he could help them in understanding the needs of the telecom market.

Product Strategy and Business Plans:

Initial Idea: The first techno-commercial plan that the Eisodus team worked out had Ethernet based access technology at its core. Recent developments in the Metro access and Metro core networks (which were becoming increasingly packet friendly) had shown that Ethernet as access technology was set to play a greater role in access networks of the future. There were several benefits attached to Ethernet such as it was well understood technology, was suitable for data related applications and was cost effective per port as well as in terms of maintenance. Eisodus team believed that with the developments of Ethernet in the core already taking place, extending the concept to the access to the first mile would complete the loop and the networks could then operate on Ethernet from end-to-end. One of the challenges in triple play service (data, voice and video) over Ethernet was to transport real time synchronous voice over the network. Eisodus planned to develop the technology for carrying synchronous Constant Bit Rate (CBR) services over asynchronous packet networks, with main focus being circuit emulation over Ethernet. This development could also help in enhancing the capabilities of VoIP services to achieve controlled jitter.

1 Name changed to maintain confidentiality This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 7 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation.

The value proposition being offered was the transfer of synchronous data such as voice and video over packet based network with efficient bandwidth utilization, Quality of Service (QoS) guarantee and service provisioning capabilities. The development was expected to take place in two stages with first being R&D oriented and the other related to establishment of manufacturing set-up. The initial plan was to develop these components either as ASIC2 or modules on network processor. Two important milestones in the development cycle were the development of synchronous packet voice processor with circuit emulation protocol and Ethernet QoS processor with flow identification characteristics.

The entire cost was estimated to be within US$ 2 million over the entire development cycle. It was proposed to use an OEM model for the selling of product as well as setting up of test sites at the client‟s premises to promote the product acceptability among the telecom companies. The above plan was finalized by December 2002. However, after a few meetings with Dinesh who was then looking after the affairs of TTML significant changes were made to the original business plan.

Idea Maturation: The second techno-commercial plan that was developed around April–May 2003 was considerably different and more detailed in terms of the product architecture and the product specifications. The main difference arose as the idea of ASIC development was abandoned in favour of developing both access and concentrator node using off the shelf components. The architecture was christened as EisoAccess architecture and it was decided to enable time critical data transfer application with QoS guarantee by a proprietary Ethernet Adaptation Layer (EAL). The basic components of the architecture were an access node which was the Multi-Dwelling Unit (MDU), connected to the access concentrator via Ethernet over DSL (see Figure 1). A network management system sitting on the top of the entire system was also to be developed. It was also decided that

2 ASIC : Application Specific Integrated Circuit This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 8 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation.

EisoAccess would provide seamless integration with Multi Protocol Label Switching (MPLS) based metro core networks. Further it was to be compliant with the emerging standards of the Ethernet in the First Mile (EFM) and IEEE 802.3ah. These standards primarily address the lower (MAC and Physical) layer issues and following them was expected to result in easier adaptation by telecom companies. Apart from the TDM quality voice service and data service with provisioning abilities it was decided to also offer both point to point and point to multipoint video conferencing at 64 kbps to 512 kbps and all this was to be done through a single standard interface unlike other available products in the market which offered different interfaces for differing services.

PSTN

Voice Ethernet Ethernet over DSL

Users Ethernet (Data, Voice, Access Access Video) Node Concentrator

Data EisoAccess Architecture Internet

Figure 1 (Source: Company Documents)

It was decided to position the product as a unique product which could help the telecom companies as well as the cable companies make optimum usage of their present infrastructure offering high speeds and also enabling smooth transition to next generation networks. In addition to this, the operational advantage was expected in the form of reduction in unutilized capacities, there by improving the revenue percentage on capital

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 9 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation. expenditure or CapEx. Additionally the target market was specified as not only the Indian market but also the Asia-Pacific market as this market had similar issues and had similar infrastructure in the last mile.

Structure of the Development Team:

Eisodus development team had Prof Karandikar as its chief advisor looking at the over all architecture and design. Initially the team worked with no clear defined roles for the members but around September 2003 formalization of roles was achieved to enable higher efficiency. The software team was led by Abhijit Gadgil and the other member of the team was Praveen Kumar. On the hardware side Zainul Charbiwala was the person in charge for complete hardware related development. Apart from the technical team, there were finance and marketing activities which were looked after by Sunil Mehta from the beginning. (For a detailed organization chart during 2005-06 refer Exhibit-1)

Prototype Development:

It was decided to develop a prototype as a proof of concept. With the basic architecture having attained a mature stage, the development team set to work. The product involved two parts hardware development for each component and then software development. Software was classified into two categories, one was to be installed on the network elements and was the embedded software kind such as protocol stack and the other was the network manager kind of software. Since the development of hardware is a long and tedious process and has to be to some extent preceded by the development of specific software it was decided to go ahead with the development of data only MDU (called the DMDU) while the other components were mostly in conceptualization or design stage. In fact it was deliberately thought that the design be kept in a state of flux to enable flexibility in case certain component of the over all architecture needed to be re- conceptualized for any reason.

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 10 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation.

DMDU Development: The first step towards the development of MDU was the choice of chipset that could cater to the technical requirements of the MDU and at the same time keep the product cost low given the nature of Indian market that happens to be very cost sensitive. This required going through the specifications of various chipsets available in the market and then procuring test boards from the parent company. However, procuring a test board or chip set itself happens to be a tedious process. The process is set in motion by contacting the respective executives of the company and then formal Non Disclosure Agreements (NDA) have to be put in place. Next one has to explain the product requirements in detail and then the chipset companies would either offer chipsets for a price or offer them on lease for development if they feel the potential product may generate substantial volumes for them. The most important component of the interaction with the chipset companies is the documentation that they give access to, which usually holds the key for the usage of such chipsets in to various applications. Often all this process takes up a lot of time as even though the document may be released but the negotiation for test sample of boards or chipsets may take longer time.

After evaluating several vendors Marvell chip was selected as it had no extra features than were required, was well understood and was being mass produced. The next step was the schematic design and layout for the development of the printed circuit board or the PCB. But Eisodus did not possess the requisite tools at that point of time so all this work was taken up in association with Telenet Systems as they had some experience in development of simple four layer boards. Since the complexity of DMDU board was not very high, Telenet could take up this work and complete it.

Parallely Abhijit Gadgil started working on the software side and the Simple network Management Protocol (SNMP) agent was written by him in the assembly language. It was one of the toughest challenges that were faced by Eisodus team and the SNMP agent was something really original that was done at Eisodus. Eisodus had decided to design their own version of SNMP agent because the commercial SNMP that were available

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 11 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation. were very expensive and the code that they were developing was well understood by the team and could be further used for future applications.

DMDU was principally ready in prototype form by end of August 2003 along with the basic documentation which was being developed simultaneously. Once the prototype was developed it was decided to test it using simulated test plans and most of these were successfully concluded by the end of September 2003. During this phase a typical problem was identified by the team which was related to rate limiting present in the chip that was being used. The development team interacted with chipset company but the chipset company could not commit on when they could offer a solution. Up gradation of the DMDU was accepted as an ongoing work by the team and it was decided that the team would evaluate other chipsets as well for the newer version of the DMDU. In March 2004, Reliance expressed interest in ADSL based MDU version and so this work was taken up to replace the existing Very High Speed DSL (VDSL) uplink of DMDU by ADSL. The design and fabrication for ADSL based DMDU was completed by September 2004.

Development of Voice based MDU (VMDU) and DAC: During the period of October 2003 to Jan 2004 several other activities were taken up. Work was started on the design of the VMDU which was one of the most important components of the EisoAccess architecture and a specific chipset was identified for the purpose. This period also marked the beginning of the conceptualization and definition phase of the Data based Access Concentrator (DAC) and the Ethernet adaptation layer (EAL). Two possible designs for DAC were identified for further evaluation and even the prospective chipsets were identified for the two designs. So, the team set to evaluate all the options for the final product. Zainul architected the final design for DAC while most of the software architecture was developed by Praveen with support from Ranoo.

As the hardware design for DAC was being developed, it was recognized that it would be too complex for Telenet to take up the schematics and layout of the PCB‟s so it was

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 12 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation. decided to look for a firm which could take up this kind of work. It was decided that Bangalore based CG-Corel would be the given the work for layout, and simulation testing of the boards that were being developed. CG-Corel had established itself as a credible player in the field of embedded system design and customized ASIC and FPGA3 design since the time they started PCB design services in 2003. Once the design for DAC was completed and all the components of DAC identified, the routing work was handed over to CG-Corel. CG-Corel completed the layout, routing and simulation testing of the complete hardware by end of April 2004. During this whole period Zainul and Balaji were in Bangalore closely working with CG-Corel team.

In February 2004 Eisodus team brought on board Dr. M. V. Pitke as an additional Director. He was an academician and veteran in the field of technology, he was himself on Board of several companies and had contacts in the telecom companies and their boards across the country. Prof. Karandikar believed that his contacts within firms could help them in showcasing their products and securing orders. In March 2004 TTML agreed to pilot test the Eisodus data solution in their network. However, in July 2004 Abhijit Gadgil resigned from Eisodus due to personal reasons and this was a set back for the development team as he had been spearheading all the software development process at Eisodus. After he left Praveen Kumar was made responsible for the software development and Zainul Charbiwala became the chief architect.

Third Techno-commercial Business Plan Released: The third version of the techno-commercial plan was formulated in September 2004. It retained most of the prior characteristics that were defined in the previous business plans but additionally brought in a few new features to be added. It named the MDU as the ENODE and defined it as the first level of aggregation. Product was slated to have a minimum of 4 data ports and 4 voice ports. It provided QoS services, TDM loop emulation and traffic management features. More importantly the uplink of ENODE was kept multi-faceted. It could have a variety of uplink through physical interfaces such as

3 FPGA : Field Programmable Gate Array This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 13 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation.

Ethernet over DSL, Ethernet over CAT 5, Ethernet over fiber etc. The basic idea was to keep ENODE versatile and be able to work optimally in any configuration of copper or fiber plant available. Another very important feature was that it was to be remotely powered through the access concentrator making its operation fairly simple. The second level of aggregation was provided by the concentrator which was named ESLAM which was a chassis based system with line cards that could be configured for different interfaces like line card with copper based fast Ethernet, line card with fiber based Ethernet, line card for Gigabit Ethernet etc. It offered features like Service Level Agreement (SLA) and bandwidth management, services like E-LAN, provided interface to TDM network through V5.2 protocol for Plain Simple Telephone Network (PSTN) connectivity and built in Metro Ethernet Forum (MEF) style circuit emulation. It was also slated to have specific algorithms to counter jitter in packets during communication. The complete architecture was remotely managed by network manager named EisoEMS based on object model and conforming to the MEF standards. It included EisoMIB to collect network statistics and usage based billing software. The plan envisaged complete EisoAccess solution to be available to the customers by mid 2005. On the marketing side it was recognized that there were different type of network providers and the solution had to be made attractive enough for them to buy the product. Also time to deployment was accepted as a very critical parameter in the entire development cycle.

Once the core solution for data was ready, it was realized that the entire solution had to be housed in an enclosure to give it a complete product form. Since the technical knowledge for such work which was related to mechanical product design was not present within Eisodus, they started to look for some company which could design and fabricate enclosures for the DMDU and DAC and identified a vendor based in Pune. The choice of vendor was based on technical competence related to mechanical design as well as ease of access as the vendor was based in Pune. Also there was urgent need for identifying a fabricator who could fabricate the given PCB designs satisfactorily. In this regard Eisodus decided to talk to Solectron and Flextronics who were well known for

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 14 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation. such work in the Indian market. Finally Solectron was selected as Flextronics was too tied with its prior commitments.

By the end of September 2004, VMDU design was completed by the Eisodus team and it was sent over for routing, testing and finally fabrication was expected by November 2004. The months of October to December 2004 saw a lot of activity. DAC/ESLAM was fabrication was completed and it was being tested. However, testing was unsuccessful as DAC was not working properly due to packet loss observed in line card. After several months of effort the packet loss problem was finally was traced to VDSL line cards and it was solved. Critical software updates were made to MDU and DAC software and the work on EisoEMS, the network manager for the entire EisoAccess product was started. Balaji played a very active role in the development of the first version of the network management system. By this time the third techno-commercial plan was released and it was decided to make the system compliant to various physical interfaces. The access concentrator boxes were designed as chassis boxes with provision for various line card adjustments. The boxes for DMDU and DAC were fabricated and now they were ready for being live tested on the TTML network. This testing was carried out beginning December 2004 and was successfully completed by March 2005 for data based solution of EisoAccess. It was decided to produce MDU in larger numbers and so quality control (QC) testing plan was formulated and automatic test suite was written for the same.

Development of Voice Access Concentrator (VAC): Two large service providers Reliance and Tata had earlier expressed their interest in the Eisodus solution. However, in November 2004 Reliance expressed its doubts about the complete product delivery by January 2005. Voice functionality or the circuit emulation to be carried out using EAL was being expressed as very important for the product to take off. Tata Communications had plans of using Eisodus solution for managed services but they too wanted the voice capability to be up and running in the product. For the implementation of voice over the EisoAccess service requirements included VMDU which was already developed, voice access concentrator and EAL protocol. Voice Access

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 15 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation.

Concentrator (VAC) was by far the most complex of the entire product so far developed and the man power required to expedite the work was not available given the funding situation (refer next section). As a result VAC hardware design and layout could not be completed by July 2005. On the software side work on call control module, resource control module and EAL transport layer and processing module was taken up and completed to some degree by August 2005. Simultaneously work was completed on the remote power unit of the ESLAM.

Other associated development was also carried out during this period. V5.2 protocol development which included developing call control and resource manager modules was done by modifying the code received from AdventNet systems. This proved to be a very difficult and elaborate task. The algorithms for jitter buffer and loss concealment too were ready by November 2005.

The period beginning Jan 2006 to till about June 2006 saw further incremental development. These included software upgrades for ESLAM, ENODE as well as the EisoEMS. Based on some feedback from marketing work was also started on complete 8 port data based ENODE with VDSL as well as fiber link and octal fiber line cards for ESLAM and this work was successfully completed by July 2006. However, the hardware development for VAC still could not be completed and after July 2006 no further development work was taken up from architectural point of view. Zainul Charbiwala resigned in July 2006 and by this time it was evident that the company was going through a major financial turmoil.

Financial Trajectory:

While the technical team was engaged in development work Sunil Mehta and Prof Karandikar were constantly on the look out for some VC funding. The period beginning 2003-04 was early days for VC„s in India and not many such funds existed. They did talk to some potential investors like Acer technology ventures, Intel capital, Jump start-up etc.

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 16 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation. but could not solicit any funds. With the development work picking up pace fund was required for the payment of salaries of the engineers and other staff as well as it was required for procuring samples of chipsets, evaluation boards and for the fabrication of PCBs. With a burn rate of almost INR 0.2 million per month they could not sustain much longer and needed more funds. SINE introduced Eisodus to SIDBI ventures and they agreed to invest up to INR 10 million in Eisodus. However, valuation of Eisodus was a difficult exercise for SIDBI as they did not have any established valuation model. For investing in Eisodus, SIDBI finally came up with a model where they proposed to take 11% equity for an infusion of INR 0.5 million. It was further decided that additional funds would be benchmarked against valuation decided by the investors during the second round of funding with appropriate equity being allocated to SIDBI during the financing round. SIDBI later changed this to discounted valuation. These terms were not acceptable to both Prof Karandikar and Sunil Mehta. Moreover, Prof. Karandikar strongly felt that SIDBI did not understand the telecom business or their product and could not bring added value on the table which could have helped Eisodus in the future. He was looking for a more knowledgeable investor. However, SIDBI and Eisodus kept running in to each other at various forums and SIDBI kept the offer alive.

Investment by Venture Capitalist: By June 2004 entire Eisodus team had been working without salary for three months continuously. While looking for funding Prof. Karandikar was introduced to Mr. Suresh Dholakia through a friend who was working for Suresh in one of his companies in the capacity of vice-President. Suresh was a US based investor and had been investing in telecom sector with a lot of success. He had earlier been instrumental in establishing Sasken Technologies which was a leading telecom solutions company based in Bangalore. Prof. Karandikar and Sunil Mehta met Suresh and found him to be interested in Eisodus and its activities. However, he wanted to invest with another investor as he did not want to go all alone. He released INR 1 million for a share of equity in Eisodus and

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 17 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation. was ready to invest up to US$ 500,000 (INR 22.5 million4) if any other VC fund was ready to invest a similar amount. The INR 1 million released by Suresh were spent on taking care of the due salaries and other procurement related costs till November 2004.

In November 2004 the only other investor Eisodus could find was SIDBI ventures and they were ready to invest US$ 250,000 (INR 11.25 million) on the valuation suggested by Suresh. Suresh came up with a tentative term sheet in November 2004 and committed to invest US$ 250,000 (an overall valuation of US$ 1.5 million with equity stake of 33% under his control). However when the actual term sheet arrived it had significant changes in it. This was unacceptable to Sunil Mehta but Eisodus was again operating with almost no funds and salaries were again due for the months of November and December 2004. So left with no choice they agreed on Suresh Dholakia‟s terms. Once the agreement with Suresh was put in place, SIDBI also started their due diligence. For continuing the operations a loan of INR 1.3 million was requested from SINE on a quarterly payment basis at 3% rate of interest. Meanwhile Suresh released INR 3 million as a part of the first installment of funds and INR 2.5 million was released by SIDBI. Once Suresh came on board he set very aggressive targets for Eisodus team and the further release of funds was contingent on the achievement of targets set by him. He also laid out the condition according to which Prof. Karandikar had to take a full time position at Eisodus by taking a sabbatical from IIT Bombay beginning January 2005. By this time Eisodus was operating with several vendors and it needed a formal team to take care of each issue, so a larger team size was required. He encouraged Prof. Karandikar to recruit people in Eisodus who had an experience of 10-12 years in the telecom market. Based on his inputs Prof. Karandikar conducted several interviews and gave about 7-8 offers to very senior level people most of whom agreed to join in as early as January-February 2005 after resigning from their present positions.

Re-structuring at VSNL:

4 Considering IUS$=INR 45 This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 18 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation.

By December 2004 the data based solution of Eisodus comprising both the access node and the concentrator had been fabricated and the testing was already taking place in live TTML network. In the same month Suresh happened to be in India at the time and got a chance to speak to Tata Tele people. His interaction with the Tata Tele people made him skeptical about the possible timing of the award of the pilot order by them. Tata Tele was by far the most promising customer for Eisodus. The possibility of the pilot order from them was one of the most important facts that had led to investment by Suresh. However, the communication business of Tata Group underwent a lot of restructuring as VSNL which they acquired in 2002 was integrated in to the overall business. The employees with TTML which was the wireline company operating in Maharashtra were transferred to VSNL. Two major business groups were carved with in VSNL; one was the Tata Indicom Business Enterprise Unit (TIBEU) and Tata Indicom Broadband Business Unit (TIBBU). Dinesh, who was a key contact for Eisodus in TTML moved to VSNL and TTML continued to exist in the form of Tata Tele subsidiary but without a focus on wireline broadband services. At the same time the enterprise business unit also moved to VSNL. This organization wide change led to changes in the power equation within the Tata Group companies and Dinesh was no longer in position to help expedite the pilot order, resulting in further delay.

Financial Turmoil and Re-negotiation with VC:

Meanwhile, recruitments had been completed and new recruits started joining by late January 2005 and February 2005. They had been promised competitive salaries with stock options but with Suresh loosing confidence in the company he did not release more funds. The INR 30 lakh that was received from Suresh was used up for the salaries and for the dues related to the demo equipment and was all over by February 2005. As a result Eisodus employees were again working without compensation. The period beginning January 2005 proved to be very problematic for Eisodus. According to Prof. Karandikar, “When we had a discussion on this issue with Suresh he said that he first wanted to get the pilot order and then only things could proceed further. The months of

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 19 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation.

Jan, Feb, March and April were very tough; we were engaged in several rounds of talking with SIDBI and VSNL also. VSNL was saying that they will give pilot orders but they kept delaying the orders.” With uncertainty in upcoming investment from Suresh Dholakia there were unresolved issues related to the level of customer support that Eisodus could provide to VSNL as well as the formal manufacturing facilities too were not commissioned.

The negotiations were again held in May-June 2005 and after long and tedious discussion documents of share purchase were signed between Suresh, SIDBI and Eisodus and there was further dilution of equity. ESOPs were also approved for the employees of the company in order to retain them with Eisodus. SIDBI also put its nominee on the Board of Eisodus for monitoring its activities. Since the incubation period of Eisodus at IIT incubation center was to be over by September 2005, it was also decided to request for extension of the same to March 2006. Eisodus was using infrastructure present in IIT Bombay such as protocol analyzer, high frequency oscilloscope etc and a shift at that point would have rendered such help difficult. After some negotiation and further off loading of equity to IIT Bombay, the stay was extended to March 2006. In the month of July 2005 it was decided to make the complete prototype that was planned, with a team of about 15 people. Finally in July 2005 Suresh agreed to release another installment of funds and when money came in Eisodus management first cleared all the past dues, and the team began work in July and by November 2005 they developed the complete product prototype (but without voice aggregator).

Pilot Order: The VSNL people gave an advanced pilot order in August 2005 and Suresh also released some money as per the agreement. In September 2005, move was initiated to get a loan from ICICI under its “Spread” program which was meant for start-up companies at low rate of interest (3-6% against the guarantee of IPR of the company) at the behest of Sunil Mehta and Suresh Dholakia. The first system was supplied to VSNL to be installed in BK

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 20 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation.

Complex in Mumbai in October 2005 after requisite testing. In November 2005 the product was officially launched. Meanwhile, ICICI and Eisodus agreed on the terms of loan which allowed loan up to INR 2.4 crore at interest rate of 6% beginning 2006. In December 2005 the price was finalized for the commercial order to be given by VSNL.

In January 2006 Prof. Karandikar had to rejoin his duty at IIT Bombay as the sabbatical period of one year had ended. In February 2006 Additional Director of Eisodus Dr Pitke suddenly resigned expressing doubts about completion of development as well as expressing doubts about the future of the company. However, this did not affect the activities at Eisodus as the team was banking heavily on the order from VSNL becoming a certainty. In order to make production more streamlined it was decided to recruit full time directors for production and operations. Corresponding interviews were conducted and offer was made to Mr. Vinod Kumar in April 2006, who had extensive experience in manufacturing related activities for the same. By May 2006 first cash (INR 2.03 lakh) was earned by the company for the pilot order given by VSNL. At this point the product was ready with its latest version, and the team was waiting for the commercial order from VSNL.

Ramp Down of Eisodus:

Although the pilot order had been received as early as August 2005 but the final commercial order which was expected to be of the order of about one million dollars (INR 45 million) was getting delayed again and again. The sales team finally reported that the issue of Eisodus boxes being used in the VSNL (earlier TTML) network was discussed at the highest level and it was decided that the box could not be used due to strategic reasons. Prof. Karandikar decided to give it another try. He contacted Mr. Kishore Chawkar, Managing Director Tata group and Mr. Srinath CEO of VSNL and requested them for appointment for discussion. Prof Karandikar knew Kishore through a friend and during the early days of Eisodus Prof Karandikar had even paid a visit to him to seek investment from Tata Group for Eisodus. However at that time Kishore has

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 21 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation. clearly said that Tata Group did not have any policy related to investment in start-up companies but they could help by giving some orders if the product developed by Eisodus was recommended by their engineering team. Since by this time all trials and live runs of the product with the VSNL network had completed, Prof. Karandikar thought that this was the time when Kishore could have helped them. But before Kishore could respond to him Mr. Srinath agreed for the appointment and during the meeting made Prof. Karandikar aware that the decision to not place an order with Eisodus was taken by due to strategic reasons as they wanted to experiment with WiMax, a new technology. He also clarified that the decision was taken for the existing time period and there was a possibility of review once the broadband demand picked up in the upcoming 6-7 months. They were however ready to give orders for some other products if Eisodus could develop those for them. Kishore Chawkar also had the same story to tell in his e-mail and so the meetings ended on a disappointing note as development of new product would have taken another 4-5 months and there was no certainty of the order.

Once this was clear that no order could be expected in the short term Prof Karandikar and Sunil Mehta informed Suresh Dholakia and SIDBI about the same. The Eisodus team realized that there were two options, first being to stay put in business, keep developing the product and also work on the products specified by Tata as their possible requirements. However, this would require further investment to take care of the salaries and other development expenses. The second option was to shut down the operations and do away with assets by putting the company up for sale. Suresh Dholakia wanted the team to continue development by retaining the team on half salaries till the order materialized. He also promised to release INR 15 lakhs of the last installment of INR 30 lakhs immediately for the same. But Prof. Karandikar did not share similar views as he had seen the problems of development in crunch of funds. He was ready to put in his effort if the investment was done in a major installment of at least a million dollars to enable proper development work. Suresh Dholakia and SIDBI were not ready to take such a risk and it was decided to close all operations of Eisodus by November 2006.

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 22 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation.

Tejas Networks Ltd. expressed their desire to acquire Eisodus with its assets, IPR as well as the team. However, after long and tedious negotiations between Tejas, SIDBI and ICICI, Tejas took back their offer. Since Tejas Networks was aware of the commercial problems being faced by Eisodus they were not keen on attaching a high value to Eisodus IPR and this was unacceptable to SIDBI and ICICI. Tejas networks however, were interested in the well trained team of Eisodus and so they made separate offers to most people personally. ICICI as a part of the SPREAD programme retained the rights of IPR of Eisodus.

Marketing Activities:

Marketing activities were started by Eisodus from day one of their inception under the leadership of Sunil Mehta. The initial focus of marketing activities was to gather information regarding the market requirements and was aimed at aiding product development. Initially only Sunil Mehta was looking at marketing and he would visit various customers for presentations and feedback, later he was joined by Prof. Karandikar as well. After investment by Suresh Dholakia in December 2005, Eisodus got another team member whose job was to constantly look out for customers in the field. So, as the company moved forward marketing activities shifted more towards trying to proactively engage customers.

Customer Interaction and Market Information Gathering: Even during the product conceptualization stage, Eisodus team worked closely with TTML to understand the requirements of the market in general. At every point of development the marketing team was in close contact with the customers and was always active in collecting the requisite information about the products in the market, and changing requirements of the customers and conveying the same to the technology development team. When the sales order from VSNL was getting delayed the team identified that the selling process to carriers was a slow process with long lead time and

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 23 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation. so they went on to identify new clients like ISP‟s who would have shorter cycle as a back up plan.

Promotional Activities and Positioning: Among the promotion activities taken up by Eisodus was setting up a website with all possible details of their products. The web site could enable the customers to understand the product their benefits as well as the possible ways to incorporate the products in a given network. Besides the web site product brochures were developed to be distributed to the customers. In 2003 Prof. Karandikar attended a conference on broadband (ICBN) held in Bangalore. The basic idea behind the conference was to present he Eisodus architecture based solution to prospective customers and possible generate some sales leads. In March 2005 the Eisodus marketing team attended the Convergence 05, a two day workshop on computing and enterprise networks held at IIT Bombay and showcased the Eisodus solution as a part of the Infineon stall. Prof. Karandikar also put up a white paper on the broadband infrastructure in August 2005, explaining the contemporary technology and upcoming technology as well as the position of Eisodus solution in the market. The idea was to position the Eisodus solution as a bridge solution that enabled the carrier to move on to the next generation technologies with minimum incremental change in infrastructure irrespective of the network composition and topology of the carrier.

Pricing: The model that Eisodus team adopted for pricing was ad-hoc. It was very difficult to evaluate the cost of software as it is mostly the development cost so Eisodus went for the hardware (Bill of Material) cost to which they added some overhead costs like marketing expenses etc. They were targeting a price such that profit before taxes and depreciation was 50% of the selling price. So, if X was the list price, they decided to sell the box at 0.6X, (with 40% discount which they may have to offer to the customers in negotiation process). It was felt that they may eventually end up selling the box at 0.6X, then 0.3X should be the profit and 0.3X should be the cost. However, it was only an ad-hoc model

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 24 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation. and the Eisodus team was not aware of models being followed by established companies for the purpose of pricing.

Eisodus team was also active on the channel development front and identified and signed an agreement with channel partners for distribution of product in Gujarat, Pune and Bangalore. However customer service was one dimension in which Eisodus could not plan its activities and had no clear policy on the issue. Another very important marketing issue is the ability of the sales force to close the deals but Eisodus sales team was not able to close the deals for their designed product. Apart from getting the pilot order from VSNL they were not able to get any customers for the product.

Technology Standards and IPR:

When Eisodus development work started Metro Ethernet Forum (MEF) and their standards were not formalized but over the years the forum was formed and they started standardizing their activities and Eisodus also started tracking it so eventually they had put this goal that they will comply with these standards. While still in development phase Prof. Karandikar insisted on the various interfaces complying with the standards specified by MEF to ensure easy acceptability of the product and also it was a signal of quality associated with the product being developed. Even the prospective investors had been insisting on such an exercise as it enabled the evaluation of the product much easily by the customers. So, the team began their study of various MEF documents by going through the documentation present on the forum website as well as other sources on the Internet. However, reading the documents was one thing and the team had to use their experience with development as well as domain knowledge to be able to work out the details and implement the final design. They redefined all the interfaces in the products with in the form specified in the MEF standards. According to Prof. Karandikar, “It really also came from the experience and domain knowledge of the people. Actually I was very fortunate to have some of the brightest people, they were really very competent. They were very good students and always knew where to look for information. It was like

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 25 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation. research to some extent. Even though they were not experienced there maturity level was matching those of the experienced people.”

Prof Karandikar had complete knowledge regarding the issues involved in filing for an IPR such as the institutions to be contacted, the way document has to be prepared etc. given his academic background and that did not prove too difficult for Eisodus and they filed for the US patent for “EAL” in 2005.

Routinization and Formalization:

Vendor Management and Cataloguing: When Eisodus team started development work with MDU it was a small thing, they were just buying components but as they went forward and started making several more components such as four different line cards in ESLAM, several versions of different components and then they had to make several boxes for the field trial so all the components had to be ordered but then this ad-hoc approach started looking unsystematic and so they wanted to systematize that. Securing a pilot order was seen as a very positive step for the company and Eisodus management started planning for putting up formal systems in place for processes. Prof. Karandikar pushed for a component database for better management of components and associated information. Eisodus soon created a database by putting in all the components into a database. So when Eisodus team was designing they first saw whether they have that component listed in their database. If that was available from 2-3 vendors, the database also listed what criteria was used to order that component apart from cost i.e. what are the other factors that needed to be considered such as availability, documentation from the parent company etc.

According to Prof Karandikar, “In India we do not have many hardware companies and that too in telecom area we have none. So we cannot really follow accepted practices from somewhere, I mean I am not aware. There may be some best practices available but we were not aware of them. All that we intended was not to follow anything but just to

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 26 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation. make our task easier and more systematic.” So, this was essentially inventory management and these practices were developed in-house through hit and trials and discussion.

Documentation: When development began operations did not have any redundancy. For example in MDU it was like one member was working on the complete hardware design, one member was working on the embedded software and other was completely working on the equipment solution. But the problem was that even if one of them resigned and left the company then the work would have suddenly stopped. Moreover if the new recruit did join he would have no way of knowing the present state of development and what had already happened etc. According to Prof. Karandikar, “See one of the things about Zainul and Praveen Kumar was that they were doing the job of three or four people, so my constant worry was that if one of them leaves the job and I recruit a new person what will I tell the person. Either I get involved personally and I know the entire code and everything which was physically impossible for me or I could push for documentation.”

Therefore, documentation was essentially source of redundancy but it also helped in putting the learning from the various development activities in a proper form and since it was read by several members before being finalized it also helped in propagating the knowledge. As a result when a new recruit would join he/she would have enough documentation to help in understanding the stage of development of the product. The above effort proved useful when Eisodus scaled up their operations in January 2005 and many people joined. With sufficient documentation having being done they were asked to read those documents. This not only created redundancy but also helped people pick up really fast and they came to understand what was happening in the company, what the product was all about and also what stage of development they were engaged in.

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 27 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation.

Exhibit-I

Organizational Chart for Eisodus (2005-06)

Board of Directors

Prof Abhay Karandikar Sunil Mehta (CTO) (CEO)

Zainul Charbiwala Asit Vadhavkar Rajesh Vakharia (Chief Architect) (Manager) (Sales Director)

Akshay Mishra Praveen Kumar (H/W Team Leader) (S/W Team Leader)

Hrishikesh Gokhale Raju Agarwal Balaji Kasal Ranoo Malhotra Atit Jain (Developer) (Developer) (Developer) (Developer) (Developer)

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 28 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation.

Exhibit-2

Magazine Report (Jan 2003) WEB LINK - http://www.expresscomputeronline.com/20030113/focus5.shtml as on 01/06/2008

Team Eisodus

Eisodus Networks — Bridging the past and the future The new kid on the block at the business incubator in the Kanwal Rekhi School of Information technology (KReSIT) is Eisodus Networks. Though a fledgling with less than two months of experience, the technologists at this start-up aspire to develop a technology for enhancing the capabilities of Ethernet-based access networks that would interface with legacy as well as next generation metro networks. Currently, Ethernet is used only for data access. But the young IITians at Eisodus envision unique innovations for Ethernet-based access networks in converged voice, video and data services. To transform this into reality Eisodus is developing semiconductor and software solutions.

Says Sunil Mehta, co-founder and chief promoter of Eisodus, “Eisodus will be a fables semiconductor company and will build a range of products in this space. The primary idea is to provide cheap access networks to the homes and offices of subscribers.” The company‟s obsession with developing a low cost access network also gave the company its name. Eisodus is Greek for „way-in‟ or „access‟.

Mehta sees huge potential for the product in the global market as currently over 90 percent of packets begin and end their journey as Ethernet frames. Owing to the technology‟s wide acceptance in enterprise networks, Ethernet seems all set to be a precursor to the broadband access market. Adds Mehta, “Offering converged services over Ethernet in access networks will be cost effective for carriers. And since our solution will enable enterprises to do this it fits well in this application space.” The company is focusing on unique ways of carrying voice and data over Ethernet in a highly

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 29 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation. optimized manner than is currently possible. The technology developed is expected to be a significant value addition to voice-over-IP as well.

The new architecture that Eisodus has developed is expected to add robustness and carrier class characteristics to access solutions. The new architecture will not only provide interfaces to legacy and new technology deployments but is also highly scalable and flexible. This, according to Mehta, will enable it in replacing these with higher bandwidth solutions in the future. The key USP is the ability to carry TDM CBR services over Ethernet, which makes Eisodus‟s solution one of its kind in the market. This solution would be able to replace ATM over xDSL, which is the transport mode in the DSL-based broadband access networks.

The marketing strategy is built in the key USP of the technology. It will offer OEMs with a technology using which they can offer the best performance and lowest cost access network solutions. The technology also has extensions for use in enterprise networking and communications.

The five-member team at Eisodus is currently focusing on proving itself and its technology in the short run. In the long run, Eisodus plans to be a leading technology provider for broadband access and metro core networks. Says Mehta, “We would like to gear up our technology to enable emerging applications in metro areas like Storage Area Networks (SAN) in a cost effective manner.” The key focus for Eisodus is broadband access networks in telecommunication networks. Beyond this the technology is also expected to come in handy in enterprise/campus LANs that are running on Ethernet today.

Unlike other start-ups at KReSIT, Eisodus has the added advantage of having academicians and industry experts on its team. This is expected to significantly reduce time-to-market. In fact, the proof of concept will be demonstrated in January. The company is also in talks with venture capitalists for the development of chips.

This casestudy has been prepared by Prageet Aeron, Doctoral student at Indian Institute of 30 Management, Ahmedabad, under the guidance of Prof. Rekha Jain, Indian Institute of Management, Ahmedabad for academic discussion only and does not represent appropriate or inappropriate handling of a business situation.