Chapter 232 Popularization and Application of Company AI

Author: unable to understand life

Chapter 232 Popularization and Application of Company AI

During this time, Wu Zhou was constantly testing the scope of use of Xiaowei and testing it.

Although, at this stage, Xiaowei has no way to directly invest in the market, because there is a high probability that it will bring about completely unpredictable changes.

However, AI assistance that matches the current stage of development, or even slightly exceeds the current stage of development, and perfectly fits the company's development will obviously be very beneficial to the company.

Wu Zhou used Xiaowei to produce a brand new company management ERP, operation assistant, scientific research assistant, etc.

After gradually introducing it into the company's operations, within a few days, there was no negative feedback from the feedback.

The efficiency is really too high.

For example, making forms and ppts before operations may require various skills.

But now as long as you describe what you need in words, or even pictures, the AI ​​system currently used by the company can automatically produce it.

And after multiple applications, each user terminal will automatically record usage habits and needs, and even directly connect to the system.

For example, Liu Bin, Liu Fei, Han Xinyan and others in the operations department used to have to do sales daily reports, inventory daily reports, product flow analysis, and industry flow change trend charts every day.

Abnormal traffic, competitive product dynamics, activity effect analysis, new product discovery, etc., various forms.

Now, it is almost a matter of turning on the computer, logging in to the system, waiting for two or three minutes, and after exporting the data from the database in the background of the platform, it will automatically produce targeted tables and even provide a preliminary analysis report.

Greatly reduces the work of the operations department.

To make a simple comparison, Han Xinyan estimated that it would take several hours to make these tables every day, mainly due to the various backends, which take time to download, and there is a lot of data. Although they have generated automatic tables, they still have to check and filter.

.

Now it only takes 3 minutes after all these forms are generated.

Comparison between one hour and three minutes.

In the past, the operations department felt that there was not enough manpower because there were too many chores. Now, suddenly, it felt that the manpower was somewhat sufficient.

As for Liu Bin, he is the general person in charge of the operations department. Every day, he opens N browsers on his computer and checks them at any time.

And now everything is integrated into the company's AI system page.

Just opening this interface is enough.

One page integrates detailed data of all online platforms.

And after using it for a few days, I even learned his analysis habits, and even produced preliminary operational suggestions based on his past operational decisions and operational effects.

The same amount of work that Liu Bin used to take a day, and it was a brain-breaking task, can now be completed directly in less than an hour, and even seems to be completed better.

"I feel like I'm about to lose my job." Some people couldn't help but think.

"If you don't want to lose your job, you need to be sensible and use your imagination. In the future, imagination is your ability."

Rational logic, AI performance is much stronger than ordinary people.

On the other side is the research center of Aiying Biology.

Cui Min is a young senior chemist at Aiying Biotechnology. Before coming to Aiying Biotechnology, she was engaged in scientific research at a university. Unfortunately, her leader did not trust her very much, and the school wanted to apply.

R&D funds require seniority and connections, which immediately extinguished most of her enthusiasm for scientific research and even gave her the idea of ​​leaving China.

And coincidentally, during that time, the headhunter from Aiying Group found her and offered her very generous salary conditions.

It is definitely impossible to say that she is not interested in the salary and remuneration. After all, she is also a human being and cannot be ascetic. Moreover, she still has a family and needs to live, and she needs money in all aspects.

Of course, what she values ​​most is scientific research, and the promise given by the Aiying Group is that it will give her a sufficient scientific research budget and complete freedom of expression. This level of trust once made her feel impossible.

Could it be that he was a liar? In the end, he finally confirmed that the Aiying Group was the most famous Chinese charity on the Internet, and visited the company in person. After the contract in black and white at the company headquarters was placed in front of him, Cui Min still remembered the look of disbelief in his eyes at that time.

Because even though he had never looked for a job, he knew that this contract was too loose and had almost no binding force on the employee.

“Is this the type of contract that Aiying Biotech recruits?”

"Of course not. Aiying Company treats different talents and applies completely different contract terms. You are an A-level talent selected by the group, so the contract terms given to you are of course the best in the company at present."

"Me? Grade A? The best? Are you sure you know my research findings? What I want to research is new drugs for rare diseases. It is not easy to generate profits?" Cui Min couldn't believe it.

"certainly."

With a skeptical attitude, Cui Min still signed. The salary was more than twenty times higher than that of the school, and the funding for scientific research projects was increased a hundred times. With such good conditions, she would be a fool not to sign.

Working for one year is almost equivalent to working in school for a lifetime.

In the subsequent work, although she only worked for less than 3 months, Cui Min was once again grateful for her decision at that time. Fortunately, she did not choose to go abroad.

The Aiying Group's original promise to her was fully fulfilled, or even exceeded.

I didn't expect that it was just a cross-border enterprise in China, but it could provide such a favorable scientific research environment.

Of course, the most important thing is that there are too many people around me who can learn and communicate with me, and almost all of them are people who love scientific research.

This kind of environment where she was surrounded by people of the same kind and could help her improve her scientific research level really made her feel like she was dreaming.

During this time, she was looking for a molecule that could inhibit an enzyme related to a hereditary disease. She tested hundreds of candidate substances, and there were three candidates, but they still did not meet her expectations and needed to continue testing.

On this day, the head of the company's information department called her into the office and introduced her to Li Yang. Li Yang was a data scientist at Aiying Group. He had been working as a data supporter in the Aiying Biodata Center department.

Work.

After Cui Min entered the door.

"Hello, my name is Li Yang. I am a data scientist at the company. The company has recently developed a new generative model that can help your scientific research project." Li Yang got straight to the point without being polite at all.

Cui Min prefers this method of communication because it saves time.

"Hello, I'm Cui Min, a senior chemist in the company. How can I help?"

Cui Min was skeptical about what Li Yang said. Of course she knew the generative model, but its reliability and accuracy were very doubtful. And from the bottom of her heart, she did not think that a generative model could replace human beings.

Expertise and intuition.

However, based on her good experience in Aiying Biological Research during this period, she did not refuse.

Li Yang then followed Cui Min to her laboratory, where Li Yang connected his laptop to her workstation.

He opened an application and displayed a very simple-looking interface with several options and parameters.

"Before we start, we need to determine the target attributes." Li Yang said. "For example, what kind of molecules do you want to find? Of course, this is more complicated. Is there anything about the binding affinity of your candidate substance?

data?”

"Yes." Cui Min said and gestured to her assistant who was not far away. Her assistant quickly started operating on the computer. After a while, a set of data was presented.

"Okay, next I will upload them to this model." Li Yang said, took out his mobile phone and directly took pictures of the relevant data, one page, two pages, three pages. While taking the picture, the relevant data information was instantly

After a while, it appeared on the computer screen one after another.

"You can check it to see if there are any errors or omissions in the data. If not, I will let him run it." Li Yang said.

Cui Min nodded, after checking for a moment.

"no problem."

After Li Yang clicked it, the interface started to run initially and entered the next page.

"Okay, now we can set the range of IC50 values ​​we want to generate." Li Yang said. He moved the two sliders on the screen until they read 0-10 nM.

Cui Min was slightly surprised by Li Yang's operation, but didn't say anything.

"Such a small value?" Assistant Cui Min said in surprise. Generally, the smaller the IC50 value, the stronger the specificity of the antibody and the better the effect. "We have not yet had a candidate substance with such a low IC50 value.

"

"You'll find out later." Li Yang said with a smile, but he was very confident. "It can explore areas of chemical space that you have not explored before."

He clicked on another option that allowed him to specify some additional constraints on the generated molecules.

"We also need to ensure that our molecules are drug-like compounds." After all, he is a data engineer hired by the Aiying Group with a high salary. He has worked here for such a long time, and he already understands a lot of common professional knowledge. So he often still

Before Cui Min could say anything, Li Yang had already opened the corresponding page.

"This means they should have reasonable molecular weights, logP values, solubility, etc. We can use some common filters like Lipinski's rule of five or QED scores to filter out undesirable compounds."

He checked some boxes on the screen and adjusted some thresholds accordingly.

"Is there anything else?" he asked Cui Min.

“Well, I need to avoid molecules that are too similar to molecules I already have,” Cui Min said. “Don’t waste time and resources testing duplicate or near-duplicate compounds.”

Li Yang nodded, and then entered 0.8 as the cutoff value for the Tanimoto similarity coefficient, which is a commonly used indicator for comparing molecular structures.

"Then it starts?" Li Yang turned around and asked again.

Cui Min hummed softly.

Then a button labeled "Generate" was clicked. The model started working, displaying a progress bar and some statistics on the screen. Alice watched nervously, wondering if anything useful would emerge.

After about 3 minutes, the model finished generating 1000 molecules that met all conditions. It ranked them according to their predicted IC50 values, from lowest to highest, and displayed them as 3D structures in a tabular format. It also highlighted

Their key features, such as functional groups, rings or heteroatoms.
To be continued...