Creating Ivideon: From Idea to Successful Implementation

Remember that feeling when you want to please a girl, but have no idea how to do this? Scroll through many options. And then you drop everything, bumping into the phrase in your head: “You will never have a second chance to make a first impression!” I remember. You try to "become better for her." Change clothes, start playing sports, sign up for dancing. But you still do not dare to approach. And then, years later, you can find out that all this was not required. She liked you so much. When I was myself. Only she didn’t understand why you just picked it up and didn’t fit.

The other day I opened a mailbox and saw a letter from the administration of the Habr. Three months have passed since we started a corporate blog, or rather paid for it. But they did not write a single article. This is largely my fault, since I considered the first article on Habré to be very important. I also believe that there is no second chance to make a first impression.

I was thinking about what to write? "We are market leaders ... the first video surveillance service over the Internet ... advanced technology ...". Sheer nonsense. I wouldn’t just read it. To hell with all this nonsense.

Many people remember the sensational project Web-elections 2012. But few people know that by the time it was launched, Ivideon servers were already on virtually all continents, except, perhaps, Africa. The number of connected cameras was already comparable to the figure that was required for installation in all polling stations. And in our technical support only 3 people worked and most of the time they “lounged”. Since our system works. All applications were processed no longer than one business day. And often not longer than an hour.

    If at the first stage we could not attract investments for a very long time, now funds are interested in us. What can I say, this is a very pleasant attention.
    I remembered how all this was born and realized that this is our first article on Habré. The history of our company. Which, first of all, begins as my own story.

Finding the perfect job

The year 2003 was in the yard, and to be honest, this is not the beginning, but the middle of the story that I could tell. But, in this case, it will turn out to be very long. Therefore, the year 2003 ...

I am a 3rd year student of an excellent university, MSTU. Bauman. Behind the heavy fights for offsets in general engineering subjects and two years of work as a freelancer at night. I was very addicted to computer games then. I was terrified to wonder how they are arranged. They read the language “C” to us. We collected simple text DOS programs by the Borland 3.1 compiler. Between them and the beautiful three-dimensional games that I played, there was a huge abyss that teachers could not fill. Therefore, in parallel, I studied game building, developing my own engine.
I constantly had my own projects, at the end of which I always learned something new. He did something for himself, paid money for something.

   By the 3rd year, I wanted to replace freelance with permanent work. Naturally a programmer. Which, on the one hand, would allow studying and its projects, on the other, it would bring money. Of course, living on a scholarship, however, as now, was not realistic. And should I tell you how many different temptations lurks in Moscow at every turn !? It was worth two scholarships to invite a girl to an ordinary theater. Therefore, I had to save and buy tickets to the puppet theater. Cheap. But how many impressions! Sitting on the front row with her! In the best places! And around schoolchildren from 8 to 12! The performance begins, and one of them takes her by the shoulder: “Aunt, aunt! You could not move ... I don’t see anything. " Well, and you, like a real gentleman, give him a rebuff ... The girl has tears of joy. She is happy to be with you! Grateful kisses simply have no end! And your date, of course, ends ... with morning coffee. Well, something I stepped aside ...

In the building of our department during my studies there were several offices. One of them belonged to a company that was engaged in video surveillance systems.
The founders of the company are graduates of MSTU. I got there with a free schedule, but in fact I devoted almost all my time to work. I really liked it. The lack of knowledge and qualifications on some issues had to be compensated by time. In the evening, at night, in the morning. Often instead of lectures. I never looked at the salary, did not pay attention to the fact that it was not even officially arranged. I always wanted to give only the best result and be proud of my work.

One of my first tasks in a new place was to send SMS notifications about movement in the field of view of the camera ... But I went much further and first suggested sending photos and then videos of the event itself. I had no idea how to do this, but it seemed to me that this should be possible. And really. At the end of 2003, the first Siemens SX1 smartphone appeared on the market, followed by the pot-bellied Nokia 6600. I should



give credit to those people who paid me a fee. They gladly allowed me to engage in this area and purchased both smartphones. And I began to study the API for Symbian and develop an application for remote viewing of video from surveillance cameras. I had enough freedom, so in fact I was not only a performer, but also the leader of this development.

$ 20 million for television on mobile

   I got tremendous pleasure from my work and from the result. In 2004, on the Symbian smartphone screen, using my GPRS application, I could remotely watch video from surveillance cameras with sound. And this was not a series of JPEG images (as many still do), but a full-fledged MPEG4 video stream. 5 to 20 frames per second! It even now seems unthinkable, and then it was just revolutionary.

We managed to port and translate the integer decoder MPEG4 and MP3. I still remember how funny the second of the MP3 sound was decoded in 8 seconds on the ARM architecture of the smartphone! And it was great to be able to do it all the same in real time.
  

A series of working days and sleepless nights dragged on. The year 2005 has come. A version was released for Symbian S60, S80 and even UIQ (with touch screen interface). And our development interested in the leadership of a large company. Perhaps, almost every resident of Russia in one way or another uses her services daily. And the deputy director of this company wanted to create a separate startup based on our technologies that would deal with television for cellular networks. He had the last Audi A8, a personal driver, a five-room apartment in an elite house ... Just yesterday he saw Putin, and the Arab sheikh, co-owner of Vodaphone, is his best friend. In general, he certainly could be trusted ... And he knew the prices perfectly. Twenty million dollars. This is the minimum for how much he promised to sell the created startup. And I, as the main developer, was motivated by a solid share of 2%)))

Vedomosti, Kommersant and many other publications published announcements of our development. There was interest from not only Russian, but also foreign mobile operators. I remember once attending a meeting at the Ararat Hotel. A rented room per day was worth more than my annual salary. It was very memorable that even the attendants there were dressed much better than me, not to mention those who came to the meeting.

Everything was wonderful. In the same year, Megaphone released its mobile television based on a Real player (RTSP streaming). But, the installation time of the session (start of viewing) was at least 45 seconds. Whereas with us this figure corresponded to only 4 seconds.

There was only one problem. Our effective manager was too wealthy and it was not interesting for him to sell our project for less than $ 20 million. And he did not find those who were ready to pay such an amount.

The last time I saw him was in mid-2005. He urgently called me and asked me to drive to the central office. On some important matter. It was a sunny summer day. And I was just walking nearby.
I was taken to his huge office overlooking the Kremlin, and there were already three people in their 40s and 50s.

With the look of a child who has just learned to ride a bike and thinks that this is a breakthrough for all of humanity, he introduced his visitors to me: “Songs” ... those very ... very famous ... They will promote us.
If I then shot that moment on video and sent it now at +100 500, it would have turned out to be a very funny release. Especially considering that they were in suits, and I was in jeans, sneakers, a NoFX T-shirt and a skateboard in my hands.

Penless suitcase

The main income of the company where I worked did not receive from development, but from the resale of equipment, installation and commissioning. And, of course, from good connections, allowing you to magically receive good contracts.

Unfortunately, for most managers, the attitude towards the development department was negative. For them, we were like a suitcase without a handle. It’s hard to carry, but it’s a pity to throw it away. In part, they were right. If my subsystem regarding the transfer of video to mobiles worked perfectly, then the monitoring system as a whole from version to version only got worse. Constant freezes and crashes led to a complete failure of the application, which should work reliably for years without reboots.



The main reason was that the head of the department, co-owner of the company, who wrote most of this system, simply did not need it. My salary was less than the market at least twice. For several years, I was promised a good percentage of sales. But what can be sales if the system does not work?
I and other developers have repeatedly suggested spending time on comprehensive testing and refinement of the product. In part, something was simply required to be thrown out and rewritten. We were ready to do it. But no one allocated time. And for large projects where our development stood, a special person was taken who traveled around all the installation objects and manually rebooted the system. Here is such a lively analogue of cron.

SaaS CCTV

By 2006, when I successfully graduated from university, I was still not on staff. Therefore, nothing prevented me from continuing to engage in freelance work. With each new order, I began to notice that everything was beginning to move actively towards the Internet.

My development was good. But, she had many shortcomings. The most important of them is working only with analog cameras connected to a capture card of our own production. Which implies the need to constantly keep the computer turned on.

I analyzed the experience of other systems and realized that their main cost for a small enterprise is maintenance. Necessities to go to the installation site and analyze malfunctions. Which often lies in the software itself.

But what if you transfer the most complex logic to data centers, and install exceptionally simple, but high-quality software on points? Then it would be possible to significantly reduce costs and improve reliability! I called this project video surveillance over the Internet. It completely fell under the SaaS (Software as a Service) model, but in 2006 I still did not know that there was such a direction. Yes, and it was not yet at everyone's ears.

I liked this idea so much that for almost a week I sat and drew structural diagrams, business models, and options for using such a system. And then he took them to the CEO of the company. He really liked the idea. He called the technical director and ... they refused her. A lot of data centers ... It's very complicated. These are years of development and nothing will come of it. Therefore, we will not do this. Moreover, nobody needs this, as the experience of “mobile television” has shown.

Our proud “Varangian” does not surrender to the enemy

   As they say in Japan, if you start to do something to achieve the goal (victory), you can lose. But if you do nothing, you have already lost.
I always preferred to send to hell those who say that nothing will work out and go forward.
   If the company is not interested in my idea, well, then I will implement it myself. But the longer I analyzed the project, the more I realized that I had nothing but an idea. Everything needs to be created from scratch. The software for working in the data center was written under Windows using non-scalable technology. That is, more than one server could not be installed. And accordingly, there could be no talk of any millions of users. The application for capturing video from cameras continued to fall and it was easier to throw it out than to try to fix it. Moreover, it was not developed by himself.

   If you write from scratch, then it is better to do it immediately under Linux. I had to actively study its features, including the network subsystem. It took me several months of sleepless nights in development and debugging, but it was terribly interesting. In the end, I wrote the first version of the network library using the epoll mechanism. And since I wanted the code to be cross-platform and used in all parts of the project, I also made a Windows implementation on the input completion ports.

Now, of course, I understand that this was fundamentally wrong. Since, from the point of view of the entire project, it was more correct not to develop a network subsystem optimized for high loads, but to create a prototype that could be optimized after launch. And so, it turned out that several months were spent and an optimized network module was received ... But why is it needed when there is not everything else?

In general, I wanted to do everything perfectly, there wasn’t so much free time, because of which it took almost 2 years to get the first well-working version. It was a cross-platform video server code (which also works on embedded platforms), applications for authorization, management, and streaming on the side of the data center, allowing you to display the video stream in Flash on the browser side. Well, in addition, a Windows application for remote viewing of cameras was made. It was the prototype of today's Ivideon Client.

I did not pursue money or commercial success. I did the project, first of all for myself, whether it is good or bad - time will tell. I did the work that I wanted and could be proud of. These criteria were not compatible with the development “for business” and were not a business. Therefore, as a result, from my point of view, a truly cool, reliable and thoughtful project has turned out.

Looking for investment

The success of VKontakte and Odnoklassnikov dizzy many. The story of the founder of Odnoklassniki was retold with particular delight when he walked around and asked for a rather modest amount of money, promising to catch Yandex in attendance in a year. Nobody believed him. Some just twisted at the temple, but he really did it. He is well done.

I understood that my project also needed investments. And without them, nothing will come of it.
The idea was very simple. Create a cloud service that allows the user to combine in your account an unlimited number of cameras scattered geographically. Moreover, it can be either Web or IP cameras connected to the application on the computer, and IP cameras for which the computer is not needed, since the necessary software for work is on board. I wanted most of the IP cameras produced in the world to already contain our service. Like DynDNS now. And, of course, all software, as well as the basic package (remote viewing of video and archive) should be free for non-commercial use. It was planned to earn money on additional services, such as SMS alerts, remote archive storage, extended support, etc., as well as commercial implementations.

I wrote letters to potential investors, went to meetings, but all to no avail. The main problem was the uniqueness of my project. At that time, he simply had no analogues. But what about us ?! Is there a successful western counterpart? So, you can invest and try to repeat in the CIS. Nobody wanted to risk on a unique project. And the idea of ​​“free” did not inspire many at all. An example was unprofitable at that time Youtube and Skype.

It's funny to say, but I even wrote to the President’s website. The fact that I have such a development that allows me to combine an unlimited number of cameras via the Internet and watch them remotely. For example, such a system would be useful for observing elections ...

And you know, I got an answer. Do not believe it - from the Ministry of Communications! It was very clearly stated there that such a project was not needed, because there is Youtube and in general the state does not need it. Given the recently spent 15 billion rubles for such a system, this sounds very funny and at the same time sad. What an irony of fate!

First money

    In 2008, I still continued to work periodically in the same place. I did not want to leave, because I was sure that in another company with a larger load, I simply could not do my project.
And as it turned out, this was done not in vain. I was contacted by one of the dealers of our system. Like, he likes how remote access works, but he doesn’t understand why everything else is so bad? How to make the system reliable?
We talked for a long time about this. And in the end I talked about my own development.

He liked everything very much and we agreed that if he finds a way to finance my project, he can count on a share. It took no more than a month, the investor was not found, but he received two contracts for the implementation of my development. From that moment we began to earn together.
Surprisingly, the system was very much in demand in companies with a branch network where remote video surveillance was required. In the same year, I first got into negotiations with him at one very large oil company. Looking ahead, I’ll say that there were many such negotiations there. And only after 2 years we were able to reach the right people and got a contract there. As they say, the main thing is not to give up.

Video Surveillance for Megaphone

With the advent of the 2008 crisis, the hope of finding investment has become much less. But, this did not prevent to continue the independent development and sale of the system. I developed, and my new partner sold. During this period, I first installed the video server on board the IP camera. It was an Axis camera. It was enough to include it in the network and she herself got in touch with the data center. Therefore, in order to connect to it, an external IP and dancing with a tambourine to redirect traffic were not required if the camera, suddenly, is also behind the router.



Following Axis, the video server was also ported aboard the Asus WL-500GPv2 router, by embedding the application first in “firmware from Oleg” and then “from enthusiasts”. It was possible to connect a USB hard drive to such a router for recording the archive and 4-5 IP cameras.



My system was for sale, but what we earned was clearly not enough to attract additional developers to the project. More precisely, the income was not regular. And I was afraid that at some point we simply would have nothing to pay wages.

We started looking for contracts that could bring us the necessary money for the development of mass service (as I called the open Internet version of the project). Meetings and negotiations with different people for six months have yielded results. They knew about us, knew that there was such a project, and one day, closer to the end of 2009, they invited us to a meeting, where they offered to participate in a tender for the development of a similar service for Megaphone.

    The company that invited to the meeting was a well-known system integrator and has already worked with Megafon on a number of projects. They promised to make every effort to win the tender, and we, in turn, to create a worthy high-quality product.

At stake was the amount that would be enough for a staff of 5 to 10 developers with good pay for a year. Since I myself was a developer, I did not want to save on specialists. Considering that quality work should be adequately rewarded. And the desire to make the first, truly massive implementation, was very great.

We were shown TK. It almost completely corresponded to our vision of the service from a technical point of view. This made us even more confident that we will play this tender. But, we were not alone. A number of different companies appeared on the list of “contestants” that never did anything like this, but they had an irresistible desire to earn money.

This did not stop us at all. And in order to have time to finalize the system for the show, I told my friend and colleague about everything, inviting him to participate. If successful, he would receive a percentage of the tender won, as well as work in a new company and twice as much salary. He agreed. In a short time, he wrote a video server interface on QT, as well as a module for working with web-cameras. Today's ivideon Server interface is very similar to that very first implementation.

We went to meetings in Moscow. We flew to St. Petersburg a couple of times (the contest was held by the Northwest Megaphone). We won two stages of the tender. Ahead was the third stage. Confidence that in the competition our development will be the best was 100%. How else? It makes no sense to work in a project that you do not believe in! All this took no less than six months. Unfortunately, almost all large companies are very bureaucratic and, as a result, not flexible.

In between meetings, I continued to work intensively and develop the system. It was not only an increase in functionality, or optimization. It was the construction and development of the service architecture as a whole. It was required to conduct development taking into account that the service will be scaled, and its individual components will be developed by the team. That is, the modules must be clearly separated and able to be tested separately from the entire complex. If today the video server supports only IP cameras, then you need to consider that tomorrow you may need to connect other sources. Now the video server writes the archive to the hard drive. Tomorrow he should do this in a remote repository.

Dismissal

Due to frequent travels and work on the project, I began to appear less often in the office. My salary was at that time 2.5 times less than the market, and my schedule (as we agreed initially) was free. At the same time, I managed to solve all my problems. Only, for some reason, the management decided that since the crisis is in the yard, then you need to work from 9 to 6, regardless of salary. And the fact that there are no promised percentages of sales is that nobody cares at all.

 In general, without going into details, I will say that in December 2009, because of this, I had a conflict with the leadership and I was offered to quit. Looking back, I can say that in this conflict I was very wrong. I wanted to create a development line within the company that produces high-quality products. This implied serious changes and a fundamental change in the approach that was used at that time. It was necessary to abandon part of the development and completely rewrite something. But, I saw a reluctance to do this and it annoyed me. Especially considering the several years of my life spent on this company. Why change something when profit is already there? I could not understand such a position. I can’t understand now. But, I forgot about the main thing. They are a guide. And they have a right to it. This is their business and their own path.



And I did it. Dismissal was a great kick for creating what is now called Ivideon. It should have been done much earlier. It was necessary to devote himself entirely to the project, without becoming angry at other tasks. But, it’s always easy to talk about this after. When you clearly see your mistakes.

Cancellation of tender

   The year 2010 has come. Year of birth of Ivideon, as a separate company. But for now, we were looking forward to the start of the third stage of the tender from Megaphone. Dates were constantly postponed. First it was January, then February. And then, and March. And then we were just told that the tender was canceled. And a completely different company was selected without a competition. One could guess the reasons for this decision.

It was extremely unpleasant for us. A lot of time was spent on this tender, but everything that is not being done is being done for the better. Looking ahead, I want to say that today, after two years of operation of this service, despite massive advertising, Megafon was unable to connect even half of the users who register with us for a month to its video surveillance.

   But about this, I will tell in the next article. Nevertheless, we found a great investor. How we created our own team and developed our project. About the appearance of competitors. What mistakes have been made and what successes have been achieved. As without significant advertising costs, Ivideon has become the most famous SaaS video surveillance service over the Internet in the CIS.