Компания Envion Software открывает девелоперский центр в Одессе

Компания Envion Software существует с 1985 года. Наши офисы находятся в Бостоне, Санкт-Петербурге, Харькове и Львове. Команда Envion Software успела создать десятки успешных проектов для своих клиентов в Америке и Европе, и продолжает работать в наиболее сложных направлениях разработки ПО.

Многие из наших проектов являются революционными в сфере IT. Например, мы первыми изобрели высокоточную автоматическую систему переноса слов, основанную на принципах морфологии и фонологии для английского языка. Позже данная система была расширена еще на 32 языка. Нашу систему используют Microsoft и Adobe, а также ряд других игроков издательской индустрии.

Также нашими пионерскими разработками являются первые десктопные издательские приложения. В них была реализована система автоматического обтекания изображения текстом, движение текста при печатании в режиме Real Time (т.е. именно так, как вы видите текст при печати сегодня).

Одним из постоянных направлений разработок нашей компании являются разнообразные лингвистические системы. Мы совершили огромный прорыв в разработке семантических сетей для английского и норвежского языков. Среди наших изобретений также stemming engine (движок нормализации слов), который поддерживает широкий ряд языков индо-европейской и семитской группы. Движок выполняет автоматический анализ склонения и спряжения слов, выдает все словоформы и нормализованные слова. Stemming engine уже давно нашел свое применение в поисковых системах и системах изучения иностранных языков.

Сегодня мы продолжаем расширять системы, направленные на изучение иностранных языков и распознавание речи, а также работаем в сфере ПО предназначенного для анализа биомедицинских исследований.

Больше информации о компании вы можете узнать:
кликнув сюда envionsoftware.com
а также здесь www.linkedin.com/company/envion-software
и здесь jobs.dou.ua/...ompanies/envion-software

Самый главный актив нашей компании — это наши сотрудники. Envion Software — это команда сильнейших программистов, обладающих выдающимися аналитическими способностями, прочными инженерными знаниями, завидной логикой и системным мышлением. Это люди, которым по-настоящему нравится работать в сфере разработок ПО.

Именно таких людей мы и ждем к себе в команду!

Ctrl + Enter

Котэ гуляющее само по себэ 03.03.2015 23:38

Наше Вам с кисточкой.
Не так чтобы очень мне это всё было нужно, но шо вы таки имеете сказать за java qa automation, с опытом, не имеющим цены?
Пришлите шо вы хочите, на посмотреть, и какой навар он с этого всего будет иметь?

Відповісти

Підтримати

осокор тауер 03.03.2015 15:26

Приятные новости, а то везде кризис и агитация уехать в восточную европу на 60% текущей ЗП:).

Tim Tretyak ASM, TS Consultant в Hewlett Packard Enterprise 06.02.2015 18:37

Ой как врать нехорошо! Особенно в месте, где это легко проверяют.

Компания Envion Software существует с 1985 года

Рыли? Пруфы в студию! А то гугл дает только ссылки 2014, а на сайте написано «10 лет».

десятки успешных проектов для своих клиентов в Америке и Европе

Список этих клиентов — военная тайна, да?

мы первыми изобрели высокоточную автоматическую систему переноса слов

Сотрудники компаний Xerox и Quark нервно хихикают, читая этот текст.

Нашу систему используют Microsoft и Adobe

Но они об этом не знают, т.к. ни одной ссылки в гугле не находится.

нашими пионерскими разработками являются первые десктопные издательские приложения.

Тут сотрудники вышеупомянутых Xerox и Quark уже заходятся в гомерическом хохоте и к хору присоединяется старичек Кнут и персонал Стенфорда.

Мы совершили огромный прорыв в разработке семантических сетей

А еще слетали на Альфа-Центавру и решили проблему расширения вселенной. Если судить по количеству пруфов написаного.
.
Давайте парируйте, а то станете тут нарицательным названием среди бодишопов.

Natalia Sheremetieva QA в The best company 03.03.2015 13:12

«If we don’t understand our tools,
then there is a danger we will
become the tool of our tools.
We think of ourselves as Google’s
customers, but really we’re its products.»
— Rebecca Goldstein.

So here is little history. I hope you folks can spare a few minutes it would take to read through this. It is somewhat longer than the average comment here.

I thought for a while before deciding to reply. An acquaintance of mine once said to me, «It is not wise to enter into a peeing competition with a skunk. In the end you both stink.» Since then I have always held that to be a very wise notion.This was a customer of ours Fred Ibrahimi, the CEO of Quark. We were chatting at a trade show sometime in the very late 80s or very early 90s. By that time my company Circle Noetic Services (d/b/a as of late 2014 Envion Software) left desktop publishing field to concentrate on computer linguistics. So he and I were no longer competitors. The product (our JV with Manhattan Graphics) I was a primary engineer for Ready,Set,Go! was sold to Esselte Letraset and I was out of the DTP game. Given that Ready,Set,Go! and Quark both paid us royalties it made it almost irrelevant as to which company does better. In any case Quark was late coming in and had to contend with PageMakers 60% and Ready,Set,Go!’s 36% market shares of the DTP market. (The other 4% were split between MacPublisher, Quark and a many other now long forgotten start-ups, some with very notable products.) Tim Gil and Fredy won that round only to be unseated by Adobe’s InDesign in recent years. Such is the life in the fast pace of the techno-universe. But then through Quark’s brilliant aggressive marketing, aggressive investment in technology, a decision to move ahead of Apple’s QuickDraw font technology limitations by creating their own screen and PostScript font rendering engines among other things, Quark better addressed the evolving needs of the market place and in the end pushed PageMaker and Ready,Set,Go! out. A few people recall PageMaker these days and even fewer remember Ready,Set,Go! although both products are still existing. Ready,Set,Go! and Sahafi, RSG’s Arabic language cousin, are still flickering in the kind hands of Diwan Ltd and we still get a smidgen of royalties from them each year.

Although I cannot take credit for that part, it belongs to Martin Rosenberg of Manhattan Graphics the inventor of RSG, RSG was the first in its linked block/box approach to building a page. This paradigm is all that is used today. Probably nobody even thinks that other ways of pasting up a page are possible. Quark XPress version 1 was so similar to RSG that it copied many keyboard shortcuts. At the time I felt that imitation was the best form of fluttery. RSG did a few other firsts, thought impossible at a time, 1. automatic on the fly hyphenation, 2. automatic real-time text reflow inside a complicated layout, 3. automatic real-time text runaround the contour of an image. Remember the computers running at 1MHz were the top of the line then. They were probably about 10,000 times slower than modern machines. Getting this much performance out of a Fat Mac required some clever engineering, especially so given that the product was coded in C and not assembler.

Our company started in the summer of 1984 by a motley crew from MIT which included my fiancé Margaret Magnus, Haj Ross (aka John Robert Ross) and myself on an idea that we could combine my talent for writing software with Margaret’s and Haj’s linguistic brilliance and solve some interesting problems. The first problem we decided to tackle was that of automatic hyphenation. We were well aware of the contemporary efforts by Knuth and his grad student Liang in solving the seemingly intractable problem of automatic English hyphenation by computer. A early copy of Liang’s Ph.D. thesis is still on my book shelf. Liang under Knuth’s direction endeavored to deal with the problem of English hyphenation and later of any language hyphenation by applying purely mathematical methods to what appeared to everyone as very random patterns of English syllables. The basic goal was to achieve the best possible accuracy given specific data size constraints. We thought we could do just as well or better using what at the time was only an uncertain insight into English phonology and the working of English stress. Anyone we talked to at the time said it was not possible and yet we did it. Our system offered a couple advantages over Liang/Knuth approach. First it did not need anyone to compile a list of hundreds of thousands of correctly hyphenated words before an algorithm could be derived. Second it offered a much greater accuracy outside the fed in corpora. And finally it ran faster. And for us it gave us the satisfaction of knowing that we cracked the uncrackable code of English stress patterns and developed a nearly universal syllabification engine. And yes we licensed this and other of our technologies to Quark, Adobe, Microsoft, Oracle, Atex, Bitstream, DRI and many others.

Coincidental with our work on hyphenation we had an idea of a product we called Understander. The idea came in part from the prior work Margaret did for NCAR in Boulder, CO where she build a compact system for accurate retrieval of relevant scientific articles form NCAR electronic database and in part from a phone conversation we had with Dick Brass, in which he suggested that we could do better than base our company on hyphenation. We should try to develop a system for personalized news delivery.

In those dark and unfriendly pre-internet days news were distributed over «wires» and only a very small fraction of the news would actually make it to print and that is all that most people would be able to know. But if you really needed to know and had the means to pay for it, yourcould for a sizable fee subscribe to the «wires» and get everything. The data would flow in a constant stream of thousands of articles. So separating the wheat from the chaff meant spending a great deal of time reading things you did not want to read. Hence the thought that if we had a tool that could with good reliability evaluate relevance we could make lives of these people easier. Still this was a small market and we thought to aim the product at the developing area of electronic scientific library materials. It was becoming more and more of a practice to require that all theses be submitted in LaTeX format. We developed a prototype that in our tests on limited corpora with our limited vocabulary and limited semantic frames set gave us 96% accuracy. So we thought that with this proof of concept we could go to VCs and raise some money. The proof of concept was doing fine on a small sample but we knew that to handle the whole language we had to do a lot more work. It is always very interesting in retrospect to look at the reactions of the people to the idea that is before its time. If Ken Olsen is famous for saying something like PCs are just a passing fad, our VCs should go down in history for saying that our search engine is not an interesting idea because we already have librarians who do this job very well. But that how the world looked to most people in 1985. Not Dick Brass, not us but to most it did.

Nonetheless we did not give up on the idea and decided that we will through licensing of our hyphenation and work on RSG finance the development of a full version of Understander ourselves. From this came the ideas of a stemmer and of the semantic net. Certainly at the time these were pretty new and interesting thoughts. It was also helpful that at the time we took interest in phenomenology and phenomenological ideas on the formation of meaning resonated in our minds and influenced how we viewed the structure of the semantic net. The notion of a meaning as a web of relationships and the relationships themselves comprising of words, with those meanings also webs inspired us. In the process of working on our net we identified over 300 meaning axis which interconnected the words in the American English language. We managed to complete the first pass through our English wordlist interconnecting words along these 300 axes before bad things started to happen in business and our lives. We ran out of money. We tried to apply for grants siting similar work that was started at Yale on the project called WordNet. Unfortunately we were not successful. In the end InOtherWords, this is how we dubbed our creation, was never fully completed and was never used commercially. The lesson we learnt that we could bite more than we could chew and we almost choked on it. Still our efforts were not completely fruitless. The project for Norwegian semantic web TROLL was started in Trondheim during our 6 months tenure there. The ideas tried there were in significant ways influenced by what we thought and accomplished already. There was a German semantic net project started based on InOtherWords too. Additionally one of the outflows of this work was Margaret’s 1000 page magnum opus on phonosemantics (en.wikipedia.org/...Sound_symbolism) that earned her a post Ph.D. professional doctorate in linguistics and also her short book on the subject for the rest of us called Gods of the Word (www.amazon.com/...Margaret Magnus or www.amazon.com/...argaret Magnus

The focus on the stemmer came from the need for rapid wordlist development for our spellchecker. Initially it was a system that only worked in one direction. It only took the normalized word form and conjugated or declined it. In time we evolved it into a bidirectional system that could both conjugate and normalize words. But here again we went against what people thought was possible. We developed a reversible system of rules flexible enough that it could be used to describe the conjugational morphological changes in a very broad set of languages. We can handle Germanic, Romance, Slavic, Persian, Indic and Semitic languages in our system. Not only our morphological framework is unique but in the process of developing the stemmer we invented a compression algorithm that enabled us to have an O(1) search time coupled with rather impressive compression results. Depending on the language an average space occupied by a word with its conjugational information in our system is from 2 to 5 bits. It’s bits not bytes.

We also had an unsuccessful fore into social media in 1999-2000. We had an idea to develop a website where people could share information about themselves or someone they loved and who’s memory they wanted to preserve, with friends and family. They could exchange messages and collaborate on joint projects. It took us a year or so but we developed it and went for the venture money. We looked for a while but the answer we got everywhere was pretty much the same. People thought that we were crazy to think that anyone would put so much personal information on the web. In four short years Facebook started. Mores change so fast these days. (I just checked and the domains we registered then are still on auto renewal and are still registered to us. They are civeterna.com and perpetualwebsites.com. The site of course is not active anymore.)

There are more projects and ideas to talk about but I think I made my point and this article is already far longer than it ought to be. So I will spare everyone anymore of my reminiscences. For a number of our ideas and initiative we did not have the business skills to bring them to the world in a big way. For some we did not envision them just right so that they could catch fire. For some we were much ahead of our time and the market could not see the value. They are not therefore on the first pages of Google search and if this was the point here then it would have mattered. But it wasn’t the point of our article and it does not matter. This article was about the culture of mind, the culture of invention and creativity that underpins our company. This is where we came from and it influences what we value and how we think, live and work together. This is what we are inviting people to join.

On the final note, when using a tool like Google one needs to take care to understand the inherent prejudices and profound limitations of the information it delivers and pay attention to it’s meaning. Otherwise one does become a tool of a tool. 2014 was a d/b/a registration for Envion Software in the USA for Circle Noetic Services. After 30 years of operating under Circle Noetic Services name in the USA I finally decided it was time to abandon my sentimental attachment to Circle and even in the US operate under the name of Envion Software as we have been operating in Russia and Ukraine for the last 14 years.

Tim, Look for the truth and reason and not for the reason to posture and insult. Look deeper than shallow Google search. Seek to forward the discourse rather than to reflexively posture and disrupt. Your life will be better off for it.

With warm regards,

Sasha Nizhnikov
CEO/President
Envion Software

PS.
By the way Quark has always licensed their hyphenation and while Xerox spinoff Microlitics had interesting spellchecking and thesaurus technology I do not believe they ventured into hyphenation space. So I am not sure why they would be laughing.

Tim Tretyak

Tim Tretyak ASM, TS Consultant в Hewlett Packard Enterprise 03.03.2015 15:12

Thanks for explanation.
It’s a lot of fakers and scammers in the Network novadays so all inaccurate info treated as unreliable by default.
Will be much better for prospective employees/clients to see short version of above text on company site or at least year-by-year company history.
.
P.S. Still not clear is Envion is successor of Circle Noetic as a company with all assets of just one of the CN founders founded it as a new private business.

P.P.S. It’s a little strange that (Xerox-Corel) Ventura Publisher is seems completely unknown for digital publishing specialist of 80-90. And yes it had hyphen feature at least from version 2.0.

P.P.P.S. As far I remember none of desktop publishing software (Quark, PageMaker, FrameMaker, Ventura) have had russian/ukrainian hyphen in 1990

Natalia Sheremetieva

Max Ischenko Founder в DOU.ua 04.02.2015 15:25

Jobs.dou.ua дает три бесплатных вакансии каждый месяц. Про Джинн уже молчу. ;)

Sergii Voloshyn Product Manager в DOU.ua 05.02.2015 09:25

jobs.dou.ua/...ware/vacancies

Max Ischenko

anonymous 04.02.2015 14:27

Natalia Sheremetieva QA в The best company 04.02.2015 16:20

Движок был сделан в 1988 на базе имеющихся ТОГДА технологий.

anonymous

anonymous 04.02.2015 16:31

Oleg Zavorin Devops в Akvelon 06.02.2015 16:03

ну разница между вашим и их подходом лишь в том что люди на это уже бабло заработали...
думаю с 1985 нормально уже ;)

anonymous 06.02.2015 16:07

Oleg Zavorin

Alexandr Szpinkowski PhD 04.02.2015 13:06

Где Вы в Одессе? Спасибо

Natalia Sheremetieva QA в The best company 04.02.2015 14:47

Да)

Alexandr Szpinkowski

Natalia Sheremetieva QA в The best company 04.02.2015 16:16

На данный момент есть только административный офис. Для команды идет подготовка офиса в ценральной части Одессы. Если вы хотите приехать к нам в гости, давайте свяжемся по [email protected]

Alexandr Szpinkowski PhD 04.02.2015 13:05

Дважды связывался с Вами по емейл, и форме feedback, но ... недельное молчание. Страница FB тоже сверхмолчалива.

Добрый день, Александр. Пришлите, пожалуйста, ваше письмо на почту [email protected]

С уважением, Наталья.

Rost Lucifer iOS Lucifer в Heaven on Earth 04.02.2015 12:36

существует с 1985 года.

Вы обпечатались, потому как в Вашем профайле Вашей компании всего 10 лет :)

И почему у Вашей компании так мало сотрудников ?

Natalia Sheremetieva QA в The best company 04.02.2015 14:54

Добрый день, Ростислав.

Компания основана в 1985 году в США. В Украине действует уже 10 лет.
А почему мало? Пока было достаточно для качественного выполнения проектов)

Сейчас идет расширение компании. Ждем сильных способных программистов в наше представительство в Одессе.

Rost Lucifer

Rost Lucifer iOS Lucifer в Heaven on Earth 04.02.2015 14:59

основана в 1985 году в США.

Обычно пишут когда было открыто представительство в Украине :)

Oleg Zavorin Devops в Akvelon 06.02.2015 16:01

обычно людей интересует сколько корневая контора работает, а не ее представительства

Vitaliy Goloborodov UX/UI Designer в Private Enterprise 21.12.2015 12:21

Феерическая чушь. Давайте тогда уже за общий возраст компании принимать время, которое она находилась по одному и тому же юридическом адресу. )))

Vitaliy Goloborodov UX/UI Designer в Private Enterprise 21.12.2015 12:18

Вы обпечатались

Кстати, Вы тоже опечатались либо просто правописанием владеете не очень хорошо. Толковый словарь Ушакова Вам в зубы. ))

Rost Lucifer iOS Lucifer в Heaven on Earth 21.12.2015 12:26

Вам бы понтов поменьше и манер дяденька добавить в зубы :)))))

Vitaliy Goloborodov

Vitaliy Goloborodov UX/UI Designer в Private Enterprise 21.12.2015 12:40

Пусть это Вас не беспокоит, юноша. Мои понты и манеры по сравнению с Вашими — просто Ваш недавний детский лепет. Лучше позаботьтесь о своей успеваемости в школе. :)

Rost Lucifer iOS Lucifer в Heaven on Earth 21.12.2015 12:46

Дяденька, Ваше хамство мне побую, как и Ваши дешевые понты :)))))))))))))))))))

Vitaliy Goloborodov UX/UI Designer в Private Enterprise 21.12.2015 13:00

Эмоциональный Вы юноша, хамить здесь в ответ на вполне, как мне кажется, справедливое замечание начали Вы (это, кстати, распространенная реакция среди не самых грамотных подростков) и, как очевидно уже по первому Вашему комментарию, у Вас рвануло таки настолько по-детски, что говорить, что Вам что-то «побую» с Вашей стороны уже поздновато. До вечера еще далеко... пишите, если что. И побольше смайлов, это добавляет Вам самоуверенности в своей абсолютной правоте, судя по их прогрессии.

P.S. Надеюсь, весь этот бессмысленный диалог кто-то потом удалит. :)

Oleksandr Chupryna Software Engineer 04.02.2015 11:24

C/C++/Qt разработчики в Одессе нужны?

Natalia Sheremetieva QA в The best company 04.02.2015 14:57

Добрый день, Александр.

Срочных вакансий по C/C++/Qt на данный момент нет.
Пришлите, пожалуйста, ваше резюме для ознакомления.
Как только понадобится сотрудник по данным технологиям, я с вами свяжусь.

Oleksandr Chupryna

Компания Envion Software открывает девелоперский центр в Одессе

29 коментарів

Підписатись на коментарі

Новини