How Artificial Intelligence and Machine Learning can be used for your project or work?


#1

Hi,

Hope you have seen the introductory video of Artificial Intelligence and Machine Learning for Managers.

In this video, we have discussed the use cases of AI and ML across various industries. Please reply to this topic and let us know how will you use the AI and ML in your work, project or vertical.

Your case study might be graded in the course and we will publish a few case studies in our blog mentioning your name.

All the best


#2

Hello all,

My team is working on innovative approaches to realize prospective and retrospective insurance cost savings. ML is currently being used on a couple of projects where we aim to build models that identify anomalies in billing and payments patterns and I would like to gain a better overview of what can be accomplished and how best to support my team.

Regards,
Svetlina


#3

Hi,

Currently there’s no plan to leverage AI at work.

In regards to project, I will like to use AI/ML to analyze data and determine certain pattern or behavior to predict the future. For e.g., based on seismographic data predict a future earthquake.

Thanks
Ajoy


#4

There are two ideas in my mind.

  1. In a Baby Monitoring system there are events recorded based on the voice and motion triggers, using AI my goal is to give a heads up to monitoring Parents/Guardian of any abnormal activity based on the past events of the baby. It can be a sudden baby’s allergic reaction when a pet is around baby’s room, or baby’s movements based on external weather conditions etc…

  2. A more futuristic AI application , is to analyze an existing patient’s voice sample from time to time, AI should be able to analyze the voice sample to detect for any probable ailments and inform caregivers.


#5

I am a Telecom professional. The telecom system can be made to achieve AI by making it understand the different actions that are taken in order to mitigate a particular error. Since action can vary depending upon the system load and available capacity the machine can learn these different scenarios and accordingly can take action at its own and thus reducing the overall system downtime by not waiting for the manual intervention.


#6

Hello,
Being from the financial sector i believe AI & ML can greatly help in analyzing and tracking Anti-Money Laundering (AML). Help the financial institutions in risk rating its customers based on the transaction history and patterns etc. Also another area that AI & ML can contribute is in processing (validating) the loan documents.


#7

I work for an IT company. The first instance that comes to mind to apply AI and ML is for resource allocation. As of now resource manager does this mapping w.r.t available positions in different verticals with the available resources and their skill sets. It relies a lot on the manual screening and separate resource manager assigned for individual skill sets. Once AI solution is in place, available positions and people can store their data w.r.t. the current and upcoming openings. System can do a prelim screening and can suggest a map of position with suitable profile.
In future it can also be used for predicting resource requirements for upcoming projects and hiring projects can be made appropriately.

Thanks
SM


#8

Hi,
I am in IT for BFS, in regards to AI/ML,

  • I will like to analyse the data to determine pattern to predict the percentage of customer buying a Mutual Fund.
  • Basis of Customer’s navigation provide better campaigns
  • Applying Automated testing for online channel to ensure the released package is of best quality

#9

Hello All,

I plan to leverage this info into clinical psychology for diagnosis of mental illnesses.
It will be supervised learning and will have training data set.

The ML project will just do what an practicing psychologist get trained initially in career during internship

regards
Sarika


#10

In the sector in which I work (air navigation servicea), AI and ML we are working on implementing AI and ML solutions to, on the one hand, perform post-operational analysis of air traffic flows; and secondly to make traffic forecasts and organize air traffic flows more efficiently.


#11

In the taxation industry, the Government can use AI to review the returns and search for anomalies, mistakes and frauds.
This will make it possible for the govt. authority to look through millions of returns and look for patterns where there are chances of frauds.
The rule based checking may not be able to see catch some trends where the returns appear correct from a broader look, however, if say trends of past returns is considered, then some anomalies can show up.
Initially, the system can be taught from examples which are correct, and which are not. Then system can keep learning and detect anomalies, which can further be reviewed and reinforced learning done that way.

This can help government to detect losses to the tune of Billions of rupees to the govt exchequer.


#12

I am going to use AI & ML in my small business where I have been converting old paper based records that were pasted into foolscap sized record books. I can see from the course description that there are concepts & ideas that I could use to enhance the conversion of these old & poor quality printed records that I have scanned using a specialised iPad & special apps. I specifically am looking forward to what I can learn so that I can “create a point of difference” that competitors won’t have & can’t match on quality & price.


#13

I’m working in health care analytics. Currently we provide sales data by using some algorithm to predict the sales for a particular region. By using AI/ML I believe we would be able to provide more accurate data which would help in better decision making


#14

in medical retail sector
It is seen in the medicine business that there is huge competition and there is tremendous pressure on the sales person on the field to achive his sales target. for achiving the sales target he starts to dump the medicine at the stockist or retailer level. due to dumping and no secondary sales happening, there is high chances that the mdicine expires at the retailer/stockist level and the company requires to replace the expired medicine.
at one hand the sales person is appreciated for the sales done but the amount of exipired medicine goes unaccounted.
so, by using AI/ML by studying the past buying data and the expired product…we can predict the order from the retailer by looking into the buying pattern of the retailer. this will help in better the production planning and also there will be very less expiry of the medicine


#15

Regarding the Telecom Industry:

  • Optimization of the routing of the network: Given the historical data in terms of amount of traffic per hour/day/month the ML can be used to predict customer behaviors and thus automatically optimize the routing and capacity of the network in order to reduce failures in the service
  • Resolution of trouble tickets: Given different scenarios of network problems related to customers, the ML and AI can learn to analyse the problem and take a decision on how to fix it

#16

Hi

Recently, we were advising a maternity hospital major on efficiency improvement and business transformation.

I realized two problems in the area of Outpatient Treatment and Emergency Diagnostics

  1. Most doctors note down the patient details in a standard prescription which is handwritten
  2. During the delivery process, all updates regarding the patients conditions are also jotted down manually

The above is creating certain challenges in storing the files and using them later because the format is shabby and unreadable and deteorates as the file becomes older.

For example, information taken during the first month of the pregnancy must be available for atleast two years. Similarly information garnered during delivery is extremely useful in further pregnancies or gynaecological issues of the patient. Doctors also opine that the above two streams of information can be used for treatment and cure in pregnancy problems.

So AI can be applied in the following ways:

  1. Handwritten prescriptions can be scanned by a system and a AI model can be built to convert the handwritten prescription into a predetermined usable format (basically digitization). The software can be trained based on the handwriting of the incumbent stuff and also the medical and prescription library available to intelligently guess the handwritten words and convert them into a usable format
  1. Statistics collected using the above information can be used to model and predict key indicators such as:
  • mother’s health
    -fetus health
  • expected due date
  • complications that may come

Thank you!