Many years later, thanks to the Party’s education, THE author got a job and happened to be a software engineer. Compared with his parents, he no longer faced the loess back to the air, but only needed to look at the vast frame and face various machine problems. Occasionally read the news, want to get rich, but unfortunately vision and courage are strength does not allow. Although we can read all kinds of English literature smoothly, we can also communicate with international first-class IT experts without barrier, and we even have to think about how to fill the gaps in the industry, so that our IT level can be on par with the real world-class. Investment of friends also have a few, generally come to ask me: this think you are reliable. Good relationship, also simply express the next view, if it is really general, general say a sentence do not understand, prevaricate in the past.

In the past year, AIOps has become a frequent word, news speculation, friends ask, all kinds of projects like well-known institutions to invest, from these information, another hot track surging market has started again!

But what is the truth behind all this?

AI is fire over the years, the AI tigers are listed, again this is close to silence AI once again appear in people’s field of vision, however, the prospectus, a large number of dubious incomes, let a person always want to know, is AI has changed the world, change yourself, or AI when tall on the AI industry turned into selling case and camera? What happened to all the high-accuracy models that were promised? What happened to all the energy that changed the world? It’s fair to say that even the market AI is still struggling to find the scene.

The suspicious income has not changed people’s enthusiasm for it. As long as AI is attached to something, it will be hot. Even the Ops field, which is the most helpless and painful and often fails to be responsible for it, is affected by this concept and suddenly becomes high and lofty.

But what is AI? Most of us think of artificial intelligence (you can imagine robots in all kinds of movies), but in the tech world, AI is probably nothing more than a technology company using deep learning algorithms.

Whether they are desperate entrepreneurs or academics, they look at the concept and think about it differently. So, the traditional operational company sweet shoulder half exposed, a tattoo is like concealed, ai scholars to succumb to the ivory tower, gold-rimmed glasses FIG leaf perplexing smile, also can not cover up its connotation and poise, successors are committing this concept came out, under the packaging of various FA, to sell the concept to investment institutions. Even in the years of more than a dozen points to buy back the bottom of the premise, there are still people happy, more than life, financing.

But what exactly is AIOps from a customer perspective?

It is not the girls who are beautiful, but the divines.

In reality, AIOps is said to be earlier than deep learning, more complete than the algorithm in textbooks, and better than the effect in papers. Such hype has indeed aroused a lot of people’s pursuit, academic halo and endorsement from prestigious universities. Therefore, major traditional operation and maintenance companies can’t wait to buy it. However, the effect is not satisfactory. Therefore, in order to deliver, we had to make one big screen after another.

However, after some time, to the so-called AIOps projects, more than 50% (may be conservative) is a pile of useless data, combined with manufacturing beautiful screen, also known as business operations, and IT is a pile of business with a pretty big led display to meet customer needs, and then put a pile of data governance open source products to stack, Moreover, we need to bring more servers (otherwise, how can we make the price higher). The focus of delivery is whether the leaders are satisfied with the large screen or not, and IT is not important whether the customers have any use in the system maintenance. The small IT operation and maintenance are still over there.

In this case AIOps is a big screen.

Finally, there are more advanced AIOps that target alarms, claim to de-noise alarms, and track root causes through alarms. Look at the various NLP and other AI learning algorithms support, looks great. However, it feels like creating another problem to solve one problem. It is hard for me to imagine how to reduce noise and improve efficiency through technologies like NLP if the alarm information of the data source is poor. But in the end, it’s a matter of scientific alarm Settings… Every time I see dozens of simple IDC alarms for the so-called AI noise reduction, the feeling is very painful…. Governance of the source is not the end of it? Of course, this kind of AIOps are not stupid, all need to help Party A to do certain implementation, what is the implementation, but also let party A alarm set a little better, otherwise the AI recognition effect is poor…

Here AIOps is an alarm suppressor.

Yeah, the original kind…

Although aiOPS from higher education are well-dressed, they do not give the magic algorithm in the legend, but they do not affect the infatuation and pursuit of capital.

On the one hand, these algorithms perform so poorly in the client field that the best students of higher education graduate to find themselves fulfilling promises that are impossible to keep. On the other hand, increased capital support, but let more people see the opportunity.

Then, from ordinary operations the manufacturer’s engineers have found that these algorithms were not so amazing, but the timing is similar to stock market index, therefore, even if there is no academic great god blessing, simple TensorFlow can achieve the same expectations, even, if your data quality is a little higher, you effect but also better.

Even if the good effect is still limited, the difference is probably equal to the promise you a small sweet, midnight to although not sun Erniang, but also full of gold foot two, angry cow lady swallow mountains and rivers.

Congratulations to the investors. Happy fairy wand.

Here AIOps is equal to the deep network algorithm.

As time goes by, the AI behemoths are suddenly worthless, and aiOps smell a chill. With the Internet questioning the four dragons, AIOPS began to protect themselves. Looking overseas, servicenow, the giant of ITSM and ITOM, is on their radar. So, AIOPS, and picked up the process of weapons, ITIL management is just needed, AIOPS and process combined, however, just the process, seems to be a little bit less hot sense, so, the most popular spicy chicken, low code to come over, together, so, another feast. Yes, you read that right, AIOPS became ITIL’s best practice, and the real ITIL, sitting alone, confused.

This is where AIOps becomes ITIL best practice.

Investors are receptive to the concept, but never ambiguous about revenue. The ANXIOUS AIOps began to think about IT seriously. The CEO discussed with the sales officer. By the way, there was a way to increase the integration income, and all kinds of IT systems could be signed. That’s what Cloud did, that’s what AI did, and now AIOps continues to do the same thing, and everyone is happy.”

It’s just that AIOps, once a vibrant, charming company, eventually became a systems integration company.

And everyone was happy. Maybe one day they’ll integrate ServiceNow+DataDog+Pagerduty as well. After all, ServiceNow+DataDog+Pagerduty is close to Alibaba’s market cap. Imagine investing in a company that can integrate ServiceNow+DataDog+Pagerduty and reap the market value of an AIOPS company. What a wonderful thing… Think of here, author oneself want to do such a company.