Huawei is essentially a family service company

Talent is not the core competitiveness of Huawei, the ability to manage talents is the core competitiveness of the company.

In 1997, a professor of the drafting group was asked in the Drafting Group: "Is it the core competitiveness of Huawei?"

Ren a total answer is unexpected:

"Talent is not the core competitiveness of Huawei, the ability to manage talents is the core competitiveness of the enterprise."

Therefore, there is a sentence in the "Huawei Basic Law" called: "Serious and responsible, management effective employees are Huawei’s largest wealth." Instead of speaking, "employees are the most valuable wealth" as many companies.

In my eyes, Huawei is more than just a company operating a communications business, but also a company that has run talent, and then supports the long-term healthy development of the business.

In essence, Huawei is a family that operates.


Core indicator

human capitalROI

Operating talents is the core of Huawei. So, how is Huawei to continue to improve the ROI (return on investment) of human capital?

I think that from the perspective of talent lifecycle management, the following three initiatives are particularly critical.

1, Bottom management of the trial period employee

The proportion of personnel who have been eliminated or transferred should not be less than 20%., Because the upper limit of the international talent recruitment recognition rate is 80%.

If you don’t stick to this basic ratio, the return on investment in corporate human capital will definitely decrease, because there will be some adverse assets to flow into the company, and the loss it caused is not an individual cost, but the high cost of organizing performance.

2Top management of employees 1-3 years

Working for 1 year to 3 years, is the key stage of the return on the return of human capital investment (Huawei is about 2 years).

Generally speaking, the marketing staff is short, the R & D personnel are long, the small businesses are short, and large enterprises are long.

The employees in the first year to the third year have to force 30% of the top, because 1 to 3 years is the highest loss rate of corporate employees. As an active capital, we must calmly face the loss, but also Do your best to reduce 30% of the top 30% of high quality capital.

How can I not let this batch of quality talents lose a lot?

Through the objective evaluation method based on value contribution, 30% of the high-quality talents are evaluated and given differential treatment and differential growth opportunities.

Here, you must dare to pull the gap and dare to give different opportunities for different people, regardless of the contest, just college degree, only the consensus.

This is an employee choice of all enterprises. The object is gathered, and people can share the excellent people in order.

3Normalization of employee exiting mechanism

The three bodies in the management of Chinese corporate talents are: coming in, it is easy to go out, it is easy to get difficult, and the orderly rotation is also difficult.

This causes the talent "slab" or "fluidity deletion". I want to give a business manager of the business manager:

Talent liquidity is more important than capital’s liquidity, because human capital is more valuable than financial capital, and only makes talent to achieve the dynamic optimal configuration of talent, thus playing the maximum effectiveness of human capital.

In any position, Huawei’s default is that it is not more than three years. After 30 years of continuous evolution, Huawei’s exit mechanism and root mechanism are running with normal institutionalization.

Therefore, Huawei can be able to enter the capability, can enter, ordered. Confused other Chinese companies can do the above three points.


"Trinity" combination

Continue to create a competitive advantage

From the perspective of the business management system of the company, ROIs that continue to improve human capital are extremely difficult things. The human management model of Huawei "Trinity" is worth learning.

1. Precision-choice – selection and configuration of talents

The biggest cost of selecting people is not the cost of recruitment, but the opportunity cost of the company. A person who is a competent important position can make things, and another uncomfortable person will give the same thing.

This is what Welch said"The ancestors"– There is no suitable person (especially leading talents), and a good strategy cannot be implemented.

Liu Chuanzhi also proposed "Take a team, fixed strategy, with a team"The three-step rules, many people ask Liu’s total "Why isn’t it a strategy, take a team, bring a team?"

Liu always said that it must be "take the team, fixed strategy, with a team", this is the matter of the ancestors. Looking for someone is "something in humanity", and finding wrong people will "things". "

Therefore, the selection of people are precise. The data of the US Manager Association shows that the average talent recognition rate of American companies is 50%.

Welch is the highest CEO we have seen. He said that he used 30 years in "Win" to increase the talent recognition rate from 50% to 80%. The talent recognition rate of Chinese companies is around 35% – only one of the three positions is selected, this is the gap.

As a manager, the accurate people are a basic skill.So, how can I quickly improve talent recognition?

Huawei began using STAR rules since 1998:


T-Task Task

A-action How to act

R-RESULT result

Description of the key behavior in the past helps us accurately determine the quality and skills of the candidate.

STAR is a structured behavior interview method. After repeated tempering, after the interviewer masters this skill, it can effectively eliminate the factors that can take a head of the head, allowing the general enterprise’s talent recognition rate to more than 60%.

In addition, companies are not only to selection of individual talents, but also learn to form the best team – make the division of labor of core talents more reasonable.

Huawei began in the 1990s"Wolf Plan"That is to achieve this purpose, the two principles of team of teamal and deputy is:The core values ??are converged, and the ability to compensation is complementary.

2Accelerate the talent – talent growth mechanism

Ten years of trees, a hundred years of tree people. Under normal circumstances, the speed of enterprises will be faster than the speed of talent growth, especially in the transformation of the company’s change.

According to statistics, the return on investment in China’s business talent training (including training) is only 10% -20% of US companies. Huawei’s leading enterprises have huge investment in talents, so we must fully consider accelerating investment benefits of promoting talents.

First, one of the things that must be done is the accelerated development of employee capacity – career planning. Now is a human society, it is an enforcement enterprise organization, so we must pay special attention to employee development.

Huawei is the first Chinese company in China to introduce the "five-level dual channel" qualification system.(Note: It is now five-level three-way channel, which has added a horizontal position – project management).

Where is the problem with low talent training efficiency?

I think the cultivation of Chinese companies in two major misunderstandings – did not do "practice teaching" and "learning".

If the learning method is not based on effective growth, it is not based on the greater value for the company, I think this kind of study is a huge investment waste.

In Huawei, basic knowledge training, case teaching and action learning is no longer the most effective model, and replaced by the "training combination of training" advocated by the non-all.

Huawei cultivated talents must have the ability to "win". Huawei globalization layout, how to copy "General" talents in front?

Huawei is a convened of those most powerful national representatives (frontline), by them, as teachers.

These teachers Take the training class as a "pre-enemy operation headquarters"Lead the students to conduct all true practices.

3, High efficiency incentive – talent long-acting power mechanism

In fact, if there is money, no money and incentive mechanisms are effective in complete different concepts.

Many entrepreneurial companies don’t have much, they can attract and motivate outstanding employees; and a lot of money, a large, unprical, incentive, has a negative effect.

For example, the yield of Huawei employee is currently Wages, bonuses, TUP allocation and virtual contoursFour parts are composed. Before 2014, Huawei was three assignments, no TUP.

Over the past decade, the implementation of virtual restrictions, staged solves the "Who Wars" – the problem of long-term interest community.

But developing to today, its side effects are getting more and more obvious – forming a huge housekeeping class, these people are lying without working hard.

What should I do at this time? The essence of TUP allocation is a deferred distribution of bonuses, which is mainly assigned to young people with excellent contributions.

In this way, the amount of "Old Eight Road" virtual equity allocation is diluted, so Huawei "is fulfilled" with the struggle ".

Since the implementation of TUP, in addition to activating some "elderly", the biggest value is to enhance the abilities of Huawei to attract and retain outstanding young employees, so that Huawei will lose human capital advantage in the Internet’s new privileges.

Huawei believes that the essence of incentives is the expected value management– The employee is not an absolute value of the amount of compensation, but the gap between the absolute value and the individual expectation value.

Greed is a basic desire of people, Huawei’s greatest contribution is Let the employee’s expectation value returns to rationality.

So, how is Huawei manage the expectations of employees?


How to manage employee expectations?

First, Huawei insists on setting challenging goals – followed by cashing in the assessmentA10% / B account for 45% / D to account for 5% Basic ratio. It is easy to achieve the goal.

The requirements for A are beyond the goal, and the goals are challenging. So it is very difficult, the more the high level is getting more, and the Huawei has nearly half of the employee is the assessment result C.

When I started the assessment, many people will ask, why do I do so good?? I sayCIt is normal.

He said it is not normal, what do I say is normal? He said A is normal, I said NO, A is abnormal, A is extraordinary.

Huawei target value is very challenging, which means that even if the company’s target reachable rate is 70% -80%, the total reward package is also very large.

On the one hand, the employees of B, and the total number of employees are around 85% of the total number of employees, they have not completed their goals; on the other hand, they find that they have a lot of bonuses.

In this way, the employee will produce when they get money."负 疚": The boss is very good to us, we don’t want to do it, the company also gives so much money, embarrassing! Must do it next year.

Confused most Chinese companies, the target value is not to say, the proportion of the assessment A is still very large, resulting in mistakes that many employees are "great", and the company will give me too little. This will contribute to the "greed" side in human nature, so that more money can not be feeding.

Summary, the pattern of Huawei’s operating talents is not necessarily suitable for all companies, but its core concepts and operation methods can give many Chinese companies.

Let Chinese enterprises make better value for "value creation – value evaluation – value allocation".

Give people pricing? Establishing a talent evaluation criterion is a premise 2021-11-29

After Huawei continues to have a "ace", he only insisted on doing one thing 2021-11-25

Huawei changed the top ten cerams 2021-11-17

Every time you look at the driving force of our advancement

[Heart News] Artificial intelligence ECG and clinical risk factors forecast atrial fibrillation

Artificial Intelligence (AI) Analysis 12 Level ECG (ECGS) helps effectively evaluate the risk of trembling (AF). However, except for clinical risk factors, it is not clear whether artificial intelligence provides meaning and further improvement of accurate prediction of AF.

The study was conducted by 12 lead electrocardiograms for long-term primary health care patients (MGH) in Massachusetts General Hospital (MGH), which was inferred by convolutional neural network ("ECG-AI"). Subsequently, three Cox proportional risk models were constructed, each model included: a) ECG-AI 5-year room fuse, b) CHARGE-AF clinical risk score, and C) ECG-AI and Charge-AF rating ("CHAI"). The researchers evaluate model performance by calculating judgment (under the area of ??ROC curve, AUROC) and calibration in an internal test set and two external test sets (Brigham and Women’s Hospital and British Bioconbound). The model is re-calibrated in the British Biological Bank, and the risk of 2 years of room tribin is estimated. The most significant electrocardiogram characteristics of ECG-AI risk predictions were studied by significant mapping, and the correlation between ECG-AI and Charge-AF linear prediction factors were evaluated.

The training set includes 4,577 (age 55 ± 17 years old, 53% women, 2171 AF incidents), including 83,162 people (age 59 ± 13 years old, 56% women, 2424 AF events). CHARGE-AF (MGH 0.802, 95% Ci 0.767-0.752, 95% Ci 0.741-0.752, 95% Ci 0.741-0.763; UK Biobank 0.732, 95% Ci 0.704-0.759) and ECG-Ai (MGH 0.823, 95% Ci 0.790- 0.856; BWH 0.747, 95% Ci 0.736-0.759; UK BIOBANK 0.705, 95% CI 0.673-0.737) Auroc comparison. The CH-AI group Auroc has the highest: MGH 0.838, 95% Ci 0.807-0.869; BWH 0.777, 95% Ci 0.766-0.788; UK BIOBANK 0.746, 95% CI 0.716-0.776). ECG-AI (MGH 0.0212; BWH 0.0129; UK Biobank 0.0035) and CH-Ai (MGH 0.012; BWH 0.0108; UK BIOBANK 0.0001) Calibration error. In significant analysis, the effect of electrocardiogram P wave on artificial intelligence model predictions is the largest. ECG-AI and Charge-AF linear predictors (Pearson R MGH 0.61, BWH 0.66, UK BIOBANK 0.41).

Table 1 Baseline Information

This study developed a depth learning model ECG-AI, using 12 lead electrocardiogram data precisely predicts the time of the atrial fibrillation. ECG-Ai uses approximately 100,000 electrocardiogram training from more than 40,000 patients from the primary health queue. CH-AI is a model of ECG-AI and Charge-AF, compared with Charge-Af, in patients with more than 80000 clinical features, a large number of clinical features of more than 80000 cases, in multiple prognostic model indicators The effect is better. ECG-AI evaluates the risk of CHARGE-AF rating containing 11 components. In addition, the study found that ECG-AI and CHARGE-AF are highly correlated, indicating that ECG-AI can largely reflect the ECG performance of the determined atrial fibrillation clinical hazardous factor. The results show that the clinical risk signal based on deep learning provides similar predictive utility, ECG-AI and clinical risk factors complement each other, enhance risk prediction skills.

Table 2 Test Concentrated atrial fuse

The Attia team has developed a depth learning model that is 80% for the model of atrial fibrillation state classification of diagnosis room fibrillation and sinus rhythm. Raghunath et al. Subsequently developed a neural network that utilizes 12 lead-optic maps to predict atrial fibrillation event. In one year, the atrial fibrillation recognition capacity is good, in subset analysis, compared with Charge-Af, performance is better. The study made substantial supplements in the aforementioned work, introduced a deep learning model, explicitly incorporating the survival time, and conducted strict epidemiological surveys, including quantitative errors, calibration, re-classification and extensive external validation. However, due to the data is not applicable, the study cannot be directly compared to the aforementioned model.

The results show that when using a strict epidemiological index to assess, the depth learning model uses an electrocardiogram to assess the risk of atrial fibrillation is reliable and effective. Specifically, the study evaluated ECG-AI of the test set consisting of independent individuals from: a) the same institution with the training set, b) independent institutions in the same medical network, c) from atrial fibrillation risk Lower different region forward-looking research queues. With the previous research results, ECG-Ai is best in the closest population closest to the training set, and the difference in different samples is small, highlighting the importance of extensive external verification in assessing clinical utility. In the end, researchers suspect that the differences in identification ability may differ from different characteristics (eg, age, baseline atrial risks), and thus the specific electrocardiogram characteristics are different from the relevance between the risks of the long-term atrial fibrillation.

This study is of great significance to the relationship between depth study of ECG hazardous signals and traditional atrial fibrillation clinical risk factors. First, the clinical hazardous factors can be performed on an electrocardiogram in a perceived manner in the deep learning model. Specifically, the ECG-AI probability and the Charge-AF score have moderate correlations in each test. In addition, the characteristics map and median waveform analysis have found that ECG-Ai probability is largely influenced by atrial decoder and a replenishment period, and the atrial structure and function may be affected by chronic diseases such as age and hypertension.

Second, the depth learning model seems to extract a risk factor complementary to clinical hazardous factors. CH-AI is a combination of CHARGE-AF and ECG-AI, which always shows better AF identification, calibration, and re-classification. This shows that the predicted effect of the model combined with the clinical risk information and the artificial intelligent veneer risk is significantly improved compared to any of the methods alone. Similarly, atrial fibrillation is significantly higher based on the risk of ECG-AI and Charge-Af, based on the use of two models alone. In addition, the predictive differences in CH-AI and Charge-AF are more significant over time, which may be related to the time accumulation effect of clinical risk factors, while the authentication of ECG-Ai remains relatively unchanged, prompting an electrocardiogram-based atrial fibrillation Risk predictions can contribute the most in short-term predictions.

The design background is explained below. First, the researchers conduct ECG-Ai training for individuals who conduct at least one ECG clinical examination. At the same time, the patient baseline information includes each component of the charge-AF rating. These two requirements have introduced potential selection bias. However, the study included EHR samples accepted longitudinal primary care may reduce bias, and the research model continues to distinguish AF in a completely independent forward-looking queue study. Second, the research training set represents an individual of a single mechanism. Training for larger samples in multiple institutions helps to generate more accurate and more universal models. Third, due to the limited follow-up of British bio-bank, the evaluation time window is 2 years, while MGH and BWH are five years.

Fourth, ECG-AI is a black box model. However, compared with the previous AF prediction model, the study uses a significant mapping and median waveform analysis, and the biological reasonable electrocardiogram changes (for example, P-wave duration prolong) have the greatest impact on the AF risk assessment. Fifth, 5 years of predictive window may represent the risk of atrial fibrillation, but it cannot be intervened immediately. However, the study also has a consistent difference in a shorter time window. It is not yet excluded that ECG-AI recognizes the previously have unexposed atrial vibration. Sixth, ECG-AI and CH-AI need to recalibrate in Biobio Bioconbanks. It is necessary to recalibrate when replacing the prognostic model. It has been very accurately estimated by simple recalibration of the incidence of the atrial fibrillation of BBBs. Tip The initial calibration error can be completely attributed to the incidence of baseline atrial fibrillation in the British branch, approximately one-third of MGH and BWH. In addition, CH-Ai has a better calibration effect than Charge-Af, although Charge-AF has a similar recalibration in Bio Biological Bank.

Figure 1 Research overview

Figure 2 identification of atrial fibrillation

Figure 3 Correction of atrial fibrillation

Figure 4 In accordance with predicted risk stratified atrial fuse

Figure 5 ECG-AI depth learning model two visual behavior forms

In summary, in three independent test sets of more than 80,000 people, ECG-AI (a depth learning model using 12 lead electrocardiogram clearly predicts the time of AF occurrence time) is compared to Charge-AF clinical risk score containing 11 components. Provide a 5-year AF risk forecast similar to the effect. The CH-A model combines Charge-Af and ECG-AI to identify, calibrate, and re-categorize more accurately. Based on the depth learning ECG’s atrial risoting assessment has the potential for wide application, it can provide accurate and promotion of atrial fuser risk assessment within a few years of ECG inspection.


Khurshid Shaan, Friedman Samuel, Reeder ChristoPher, et al. Electrocardiogram-based deep Learning and Clinical Risk Fabrtors to Predict Atries, 2021, In Press.