Transformation Ai, Gen Y Huawei Java Engineer's story

Written in front: JAVA is the best language in the world; the only unchanging thing in life is change; from the inside Out is the new, the outward is broken.

One: Xi ' an and Huawei
My name is Qin Feng (pseudonym). From the time of student contact Java to write the first line "Hello World" has been more than 10 years, 2011 formally joined the Huawei XI ' an research institute has been responsible for access network, automated testing tools, distributed scheduling system and other projects. Here to brag about: people who can enter Huawei are cattle, so that there is a basis, we took over the system is mainly used for distributed execution of automated test scripts, initially only more than 10 PCs, the interface is very simple, each time only one project team to run thousands of test scripts, And the expected demand is too poor to support the dispatch of the Department, the reasons and the former left behind the pit still can not help but lament. The same year July catch JavaSE7 released, our six people decided to give up the way of PC testing, took the way of cloud virtual machine deployment, from 50 virtual machines to 500 virtual machine debugging operations, after nearly 2 years of exploration attempt, finally our distributed automated test script, Support for the entire department of Super 10w test script scheduling execution. It took a night to get out, and the back was optimized to 2 hours to run. This increase in efficiency is the power of programming, this change is the Java light deployment easy to invoke the change, Java is the best language in the world!

The success of the project made me look forward to it, so that I could soon be promoted to a more senior position: Senior Software engineer. The title of Dora Feng! From then on, talking about my career, I can say, "Yes, I am a senior software engineer, working for Huawei." "The light of the village is there." When I worked in the same post for two years and still had no promotion, I realized the reality.

My job is mainly to be responsible for an old automated test system. It has been in maintenance mode for many years, but has exceeded the load due to increased load. It will often die silently, or output the wrong information. I need to spend a few days analyzing and finding out why, because no one has written any documents except for the original requirements. None of the work I have done can be measured in numbers. Or there is no visual representation of what is being done.

After working in Huawei for 4 years, I chose to leave, many people did not understand that at that time, Huawei was in a rapid development stage, a lot of promotion opportunities and generous welfare benefits. But none of this can give me a sense of security, perhaps confirming the phrase that the only constant in the world is change. Every day nine to five of life, the constant work of the content, bring people is not a sense of security. In such a fast-growing era, the cycle of disruptive change is getting shorter. Stability is a dangerous signal, always bring people inexplicable anxiety. How to break through their own, how to keep up with the pace of the times, become almost every day will involuntarily ask their own questions.

Second, Beijing and the smart point
One day in the circle of friends to see a tweet about AI, "using artificial intelligence instead of repetitive work", the first reaction is that this is not the direction I want to find (I actually want to use artificial intelligence robot to replace their work/laugh cry). After a burst of joy, soon calmed down, 30 years old, programmers feel older, let alone to transform into a strange field.

Not to mention the wife, all the factors do not allow me to make a similar crazy decision. That period is my self-examination most of a period of time, once a little depressed life, everyone has a lot of problems to overcome, to solve, in the face of difficulties we choose to give up, or to meet the difficult depends on our own. Depression can not solve the problem, fear can not change the status quo, only positive efforts forward is the most correct choice.

At home during the period of time to start learning algorithms to contact artificial intelligence, while retrieving the nearly deserted math, English, the other side efforts to supplement new knowledge. This has persisted for 5 months, using the time available to replenish a lot of "energy". The new question is coming again, it is impossible to continue such unthinking, not to do scientific research, nor to have the opportunity. It is only the right way to use the actual scene.

Chance coincidence met Dr. Hu (Smart Little founder Hu Yunhua, usually I call "Huber", he was in XI ' an artificial intelligence lectures), after several conversations, decided to give me a chance to try. Since then I formally contacted the intelligence point, entered the this AI technology-driven innovation enterprise, to the company on the first day to listen to Huber shared the "artificial Intelligence era algorithm engineer how to Grow", harvest or a lot of, master a door even a few programming languages, mathematics, English is the basis. Then self-study accounted for 10%, exchange discussion accounted for 20%, the practice accounted for 70% of the work. Experience the deepest is full of thought learning machine learning can solve a lot of problems, sometimes in the actual work or a blank head. Also have specific problems specific analysis, such as the calculation of semantic similarity, see a lot of computational errors, may be found to be typos, perhaps the problem of participle, there may be less domain keywords. Model has many why choose linear model, do not choose decision tree, depth model must be suitable for your current project? So the algorithm is meaningful in combination with practice.

Inf. The first is to follow the video course, starting with a simple foundation, and some techniques may need to be learned more than once. The first time may be clear a few concepts, the second time to understand the principles, the third can be started to do a few simple experiments. In the attention to rapid learning, you must not neglect to become their own, or look at more than just stay on the surface.

Modest, rashness is a phrase that is not obsolete at any time. Around for several years of colleagues, CTO Moyu (here to focus on, we all affectionately called him "great", because he is not only "the beauty of programming", or deep learning and tense flow framework editor, before Bing later also did a humming to recognize songs algorithm), Huber the books and paper they have read are not at an order of magnitude, and they are continuing to work. It is a happy thing to feel able to work with them and to be diligent in this field.

Huber from the founding of the company, has been thinking how to bring people and cultivate people, everyone said he "assuming". In fact, it is closely related to his experience in Microsoft Research Asia bringing interns, Ali band team to develop technical talents. So he put forward the Touch AI talent Training program, with the resources of the previous academia, please come to Peking University, Tsinghua University, the Chinese Academy of Sciences furcated Hoof lectures, so that everyone theory with practice, hard also looked up the road. Of course, there are the CTO for me to tailor the study plan, a bit drops, extremely moved. Although, this is not a matter of overnight, but really let me in the direction of professional to full of confidence and hope.

I also often think about life, if one day to look back, not because of wasting time and remorse, not because of mediocrity and shame, for the dream and strive hard, even if ultimately not as you wish, but enjoy the process, but also worthwhile.

This is a decade of the transformation of the Java Engineer story, I am not good at lyric, so I do not expect it wonderful, but I want to share my transformation of the story to everyone, I know there are a group of people like me, perhaps at this moment they are also thinking I have thought or tangled problems. It would be a great honor for me if my transformation process helped and inspired you.

Transformation Ai, Gen Y Huawei Java Engineer's story