Does One Size Fit All? -Can AI Really Fix A Broken Process?

7 minute read

Can AI Really Fix A Broken Process?

let’s explore this question with a scenario…

The Scenario

Art work from Classromclipart.com

Mark has had a toothache for the three days, he’s tried to ignore it but it’s now affecting his performance at work. He needs to see a dentist but he hasn’t been for a few good years and so he will need to register with a new dental clinic. He calls clinic and finds out that he must be registered with the clinic before being seen for a standard or emergency appointment. So the clerk emails Mark a set of forms to fill in which he can either send back via post or hand them in at reception on the day of his appointment.

Art work from dreamstime.com

Mark takes the forms and documents with him to his appointment. The clerk spends a few minutes entering the information. After which Mark attends and completes his dental appointment. A few months later Mark gets a new job on the outskirts of town and moves closer to his new place of work. He calls his dentist letting them know his change of address. The clinic send him a copy of his old information to update and mail for updating.

The Problem

While Mark is one case were patient details must be changed the clinic has thousands patients will call on a daily basis to update their details on the dental system. The current patient dental system is setup in such a way that entering or updating patient information is a time consuming and laborious job. So much so that there is a backlog data to enter onto the system and the manager of the dental clinic has to pay clerks weekend and overtime to try and make a dent in the backlog and it’s proving to be expensive on all fronts.

clerks overwhelmed with paper work

The Solution

light bulb moment

The CEO and manager of the dental clinics put their heads together and decide that they want to utilize the power of AI in their practice. In particular they want to use a computer vision system to recognise handwritten text so they can automatically scan the text to the dental database. In this wise patients can hand in their updated details and all the clerks have to do is scan the documents through the computer. The dental company is willing to spend pretty penny in order to purchase, configure and integrate such an AI system with their existing computers as they have calculated that it will save them money in the long term.

The Result

After the upgrade the patient process still remains the same but now instead of a person entering the data; the system automatically scans and recognises that text and puts it in the computer. The managers celebrate the success of the work. money well spent.

A Closer Look

If we step back and think about the process more holistically you may realise that the process should have been fixed in the first place before considering an AI solution. Essentially nothing really changed except for one step which is now AI enabled. Instead of employing an AI system to solve their problem the dental company could have given patients a unique patient URL that would enable each patient to update their own personal information. If done correctly this would not only save the company money, time and resources but it could also be used to strengthening existing data protection policies which AI alone can not do.

The Moral of this Story

As seen from the example given their problem could have been solved without the complexity of AI or in this case computer vision. The moral of the story here is not to blindly follow the current trends in technology without thinking about what problem your trying to solve. Your company’s digital journey is important. And as such all process should be digitally optimised before reaching for AI technologies. Instead we have today are many companies and orgnaisations jumping on the machine learning bandwagon if only to seem hip and trendy when in reality they really don’t need to.

A Shining Example

With a population of 1.3 million people, Estonian economy is one of the biggest and most surprising success stories in the last 25 years. Estonia has the most technologically-savvy government in world history. In Estonia citizens have an ID card connected to a series of databases which contain each citizens medical history, financial records and a bunch of other personal data. The technological efficiency in which Estonia’s government operates provides massive relief from the pains of bureaucracy. When a child is born information about the birth is sent directly from the hospital to the population register this prevents the creation of excessive paperwork and saves time. Such a digital system will allow citizens to share any piece of their digital identity in an instant such as a mortgage lender or if you’re applying for a job all you have to do is just hand over your ID type in a pin they’ve got all the information they need. Doctors working emergency that have instant access to a patient’s full medical history would be able to better life and death decisions. Without this readily accessible information even an AI system will not be able to make any kind of accurate decision that a doctors in the field can trust. In order for to be used effectively it can not be deployed in isolation. Instead AI needs a constant flow of information in order for it to continually adapt to new situations. For this to occur a fully optimised digital processes and systems is essential. In the case of business and governments this means first moving from and optimising legacy processes and systems before trying to solve problems with AI first.

Even this is not enough

Not only do business and governments need to rethink their approach to AI and upgrading their legacy systems and processes; they must also focus on the up skill of the nation of people as a whole. Today many business and governments are willing pay top dollar to data scientist and machine learning engineers in order to become global leaders in AI; yet more so than ever these job vacancies remain vacant. What happened to all the highly qualified individuals who could take up the mantle of data scientist and machine learning engineer? In addition to this governments are formulating policies and allocating funding to build AI institutes, to attract more students to complete PhDs in machine learning. These approaches are laudable but inadvertently put the cart before the horse.

Unfortunately there is a massive shortage of data scientists today; training has not caught up to the demand –and the talent pool for AI expertise is pretty shallow. In addition there is often no consideration extend to the need for product managers, operations teams and business strategists who understand how and when AI should be used. However despite the global shortage of qualified data scientist and machine learning engineers Google, Facebook, Amazon and Uber all seem to have no problem recruiting data scientist and machine learning engineers and making great leaps in AI and digital technologies where other business and governments fall shot. The reason has less to do with these companies offering a better salary than their competitors and more to do with their approach to finding and hiring talent. Their solution is to hire your average developer, analysts and business strategists then simply train them, teach them exactly how to use machine learning techniques instead. As such not only can these companies take advantage of the naïve of creativity and all the mental energy that it brings it also instills a sense of being valued by their employers who seemly go out of there way to train them instead of expecting them to be that unicorn data scientist. Moreover in Estonia children from as early as preschool are introduced to basic concepts of programming, logic and coding in attempt to build a nation of people are digitally literate and have smarter relationships with technology, computers and the web.

Many AI policies issued in the by governments in the last two years are primarily concerned with becoming global leaders in AI technology with little consideration given to building the types of digital infrastructures that can support AI or focusing on the general upskilling of their nation. It is if for this reason that digitally enabled countries like Estonia and companies like Google and Facebook will dominate the AI industry.

Hope this helps