In the era when data can’t mix into it disdain chain
He decided to all in his life
Science and technology maker: what do you think is the first important node in your life?
Guo Wei (hereinafter referred to as Guo): it was my junior year. I have been exposed to code and programming since I was a computer in primary school. When I got to university, I began to think about the skills I learned. What can I do in the future? Language, code, from basic, Pascal, C language, Java It’s always changing, and there will be new languages in the future. What I’m thinking is,What is unchangeable in IT technology?At that time, I tried many directions. There were various laboratories in the University, such as network laboratory, database laboratory, Architecture Laboratory for chip research, and so onOnly data remains unchanged。 So I decided to join the database laboratory in my junior year. My graduation thesis is about data mining.
What was your initial understanding of data?
Guo: at that time, databases, data source groups, and data relations were the only ones that could be reached. In fact, any objective existence could be abstracted into data. Language and technology were constantly changing, but the data was relatively unchanged, which made me think it was a long-term thing to do. Although I didn’t have a high level of philosophical understanding at that time, I was fascinated by data, and graduate students did not hesitate to choose data related majors.
Department: in 2002, there were graduate students majoring in data?
Guo: really not (laughter). At that time, it was also called information and signal processing. In fact, it was data mining and pattern recognition. The earliest data mining was called data miner, which was called data miner in Chinese. When you go out to chat with people, people always ask what you dig for? The explanation is quite troublesome. It can only be said that it is not a coal digger.
Department: in those days, the data specialty was regarded as the IT circle that despised the end of the chain?
Guo: you can’t get into the chain of disdain at all (laughter). Even in the IT industry, we don’t know what we do, “do data? Oracle？” It’s all this reaction. The University participated in the largest data warehouse project in China, which was planned and constructed by China Mobile. I participated in the system specification compilation of China mobile business analysis system, including 1.0, 1.2 and 1.5. This is the first Bi and data mining project in China,My graduation thesis “customer churn analysis of process neural network” is verified with the actual data of Shandong Mobile, the effect is better than SAS customer churn analysisI got an excellent paper.
In retrospect, the data work was greatly restricted by data collection at that time. Now there are intelligent IOT devices and computer vision such as face recognition. At that time, we can only do some things based on the traditional data warehouse, such as what consumers buy by card. But it’s interesting enough that I’ve had a belief that I’ve mentioned on many occasions,“Data has a soul, and I will spend my life pursuing it.”.
Add Teradata for beer diapers
Choose to stick to faith in the industry trough
What are your criteria for choosing a job after graduation?
Guo: the creator of the word data mining isHan JiaweiTeacher, in his class, he came into contact with the case of beer diaper (Note 1), and thought it was too interesting. After graduation, he went to Teradata, a big data company that created the beer diaper case, as a data architect, and participated in the construction of Inner Mongolia Mobile Data Warehouse as a project manager. Working in Inner Mongolia during that time was impressive. In winter, when it was 30-40 degrees below zero, the sand could scrape blood from the face. When going out, you had to be fully armed from head to foot.
Ke: it sounds like a lot harder than the general it practitioners. Have you ever thought about giving up?
Guo: in fact, many of the students and colleagues who did data work changed their careers, but it was not because of hardship. The year 2000-2005 is a hot peak in the data industry. Bi and data mining are hot spots,However, some enterprises found that the closed loop of Bi is relatively long after investing a lot of money, which is not perfect. As a result, the whole data industry has gradually turned from hot to cold, and the income and rising space of practitioners have been affected. I really rely on my faith to stick to it. From Teradata to IBM, these are Bi and data mining professional manufacturers. But at that time, it was obvious that because of the reduced investment, the iteration speed was slowed down, and a lot of work was repetitive, inefficient and lack of innovation. It’s hard to make a breakthrough in this kind of work. After thinking about it, I went to CICC.
When did you start to participate in the management work?
Guo: during Teradata, I tried PM for the first time. After all, my technical ability has a ceiling, and I can’t make sure about some things. At first, I thought that management was about design, but I found thatThe most important management skill is communicationAt that time, I suffered a big loss. The project itself was not difficult, but after doing it, I found thatThe person who asks for the demand is not necessarily the real end userAccording to the requirements of the project is the real user of pain. I realized at that time,Project management must penetrate the phenomenon and see the essence of demand. Also, I strongly recommend that you learn PMP.
A good wife can’t cook countless meals
Push forward the project of 100 million yuan, tired and happy
Why did you choose to go to CICC?
Guo: because many new technologies, especially data technology, will take the lead in the application of investment banks. It is closest to money, and data and money have a natural fate.
Section: from data professional company to enterprise, do you have any trouble in adapting?
Guo: when working in professional enterprises like Teradata and IBM, the project and scope of work are defined. You just need to do it. But it’s not the same in enterprises. When I went to CICC, I found that the situation was two extremes: in the part of transaction, a gifted young doctor from Cambridge took the model analysis with him, which was very strong; but there was nothing in the internal management,So my first job in CICC is ERP. I have to get the data in hand.Until now, I’m quite familiar with accounts receivable, accounts payable and manpower (laughter). At that time, many investment banks didn’t complete the electronization. To be honest, people didn’t really need it. The list + excel was basically able to do it. At first, I thought it was a little backward, butNow I think that there is no way to win, excel is still a very powerful tool.
What work have you done in Sinochem?
Guo: first, we should do data regulation, and then we will do historical analysis and calculate risk exposure. We started to make a very advanced thing in 2010,Network finance, the entire system has been made, online account opening, online trading I didn’t expect that the compliance review failed.It took a long time, and the results of more than a year’s work were idle there, and the team didn’t know what to do. Wanda provided a very good opportunity. Wanda has a lot of formats and data. I think it will be fun.
K: what can Wanda’s business data do?
Guo: WiFi Positioning is to use smart POS card swiping information, parking information, and intelligent WiFi to record the moving track and other data. After that, all the data are connected to analyze the behavior of consumers in Wanda Plaza, so as to optimize the service.
Wanda is really rich. With 103 million yuan in the whole project and more than 20 optical manufacturers have recruited, they are the top suppliers in the world: Cisco, Aruba, moto Circle in Dalian Wanda Plaza office closed development.Wanda WiFi Positioning projects are leading in the world. These suppliers have technology but no ready-made products, and their engineers are excitedFor more than three months, he has made a three-point positioning system.
This project has gained a lot for me. Not only did I start to gain a little fame in the circle, but also made me realize that data has evolved from online plot flow to offline behavior analysis flow, with a much broader vision.
Ke: what kind of experience is it to play an important role in a project with a scale of 100 million yuan?
Guo: Wanda’s management is very strict. When it’s time to hand in the results, it often gets off the plane and goes to the scene for testing. It’s busy until dawn the next day. The advantages of strict management are obvious,At first, we tested four Wanda squares. After success, it took only half a year for all Wanda squares in China to go online. This efficiency is terrible. Of course, it is also important to have a sufficient budget。**
Also encountered problems, or communication, let me more deeply realizeCommunication isThe top priority of management.Although Wanda is promoting the project from top to bottom, I still need to explain to the responsible persons of various departments and squares what this is and why, otherwise, the cooperation of the other party will be poor, and they will always ask why they want to give you the data?
The biggest dilemma is that the heart is too soft
Management = the right of the two is the least
Ke: do you feel like a fish in the water when you get married with Yiguan and the data experts join the old data company?
Guo: Mr. Yu Yang, the founder of Yiguan, hopes that I can do product and technology empowerment. As an old data company, Yiguan has been able to do data for more than ten years. I have great respect and decided to give it a try.
When I arrived here, I found out why entrepreneurship is a life of death. The entrepreneurial curve is very high at the beginning. Everyone thinks that there is a lot to be done. In the process of promotion, we gradually find that the reality is cruel. This is not the end. If we go further, there must be a deeper pit waiting for us. When I first came to Yiguan, there was already a team of more than 20 people here. My heart said that it was good. The guys were very cute. The earliest project was Yiguan Qianfan, which analyzed the data of mobile app. The main technical function was to collect data. So do it. As a result, there are many bugs. The database is often hung up in the middle of the night, and there are some problems that can’t be explained clearly. At that time, I started to stay up all night, and my hard work team was very hard. The boss make complaints about it.
As a result, there was a major accident. The data congestion at the receiving end led to an avalanche. The team changed the situation for three days and failed to solve it. I remember very clearly that my boss came to me on my birthday in 2016 and didn’t swear,Say “Happy Birthday” and then “thank you for your work.”(laughter). It’s hard. It’s my birthday, but I’m excited,I never changed my job because I couldn’t solve technical problemsWe must solve the problems before we go. After careful consideration and deliberation,I came up with the control and calculation logic of Yiguan big data “cloud + terminal”. Now I have learned the technical language Lua used at that time, and used a piece of code one day to realize the prototype and onlineThe world is quiet and the problem is solved.
Coe: at the same time, I feel that the CTO code tapping painting style is desolate
Guo: CTO, in terms of technology, if you can’t make a decision at the end of the day, you have to go to it, whether it’s writing code or getting technical users, because there’s no one left behind. You’re the last insurance for the company’s technology. After this, a good friend in the technical circle said to me, “don’t be soft hearted. The whole team has been optimized. You can’t fill a hole as fast as they can dig a hole. I think so. I can’t do it without making up my mind.”.Later, when I was sharing, I said, “to be a CTO, you have to be a bit vicious.”In fact, this is the lesson learned.
Many friends asked me what the focus of CTO’s work was. I said that the most important job was to find people,Just like a girlGood HR personnel may not be recruited and must be directly employed by the boss (not embedded advertising).CTOstillTo be responsible for the establishment of a cultural system, it is impossible for a team of more than 100 people to know what each employee is doing. You need to use culture and values to guide and correct.
Coe: it seems that the difficult problem made you complete a spiritual transformation?
Guo: after that, there are also challenges, but they are all predictable. If there is a pit, fill it in early, if there is thunder, it will be arranged early. The challenge of the new product line and the entanglement of whether it is open source or not There will always be pressure and difficulties, but I still have a lot of control over things.
To be a CTO, you need to be able to withstand pressure. There are not many people in the company who know the technology. Many problems can be thrown at the technical team. As the leader of the technical team, you have no way to go back. You are brother paoze. You must be able to hold on and handle it well. Most of the time, the problem that technology managers face is not the priority of two interests, butTwo tuoxiang, one big, one bigger. If you don’t eat, you should double it immediately. Sooner or later, you must eat itAfter eating, they should also smile and convey positive energy to the team, so that everyone can feel at ease.
Now there are more than 100 people in the team. Together with the team, they successfully entered the Apache foundation for their open-source project dolphin scheduler. Yiguan has become a member of batjThe only company in China that is not a major open source project has been selected by Apache fund.
Thank you for your patience
If you like Guo WeiThe story,
Welcome to share with your friends.