My Summer Internship Journey with Ismile Technologies
One of the most essential things of doing Masters in the USA is a Summer Internship. My internship landed in the most forgettable phase of Summer 2020 with the unprecedented Corona Virus. I was selected for the Data Science Intern role at Ismile Technologies. The company is in its early stage and has branches in the USA, Canada, London, and India. The company is a service as well as a product-based company working on many recent and top-notch technologies such as advanced analytics, big data, cloud, databases, DevOps, and infrastructure management.
What did I actually work on?
I was a part of the Enterprise Data Science team which was developing products around Artificial Intelligence/ Machine Learning domain. I was a part of two teams – Ecommerce Bots and Hospital IOT.
I was a team lead for Hospital IOT product, this is a project under the Health Care and Life Science domain. There is no doubt that this industry needs a whole lot of attention than it gets. We are planning to automate and reduce the resources associated with the patients. Every patient needs the attention of at least one nurse/one caretaker. To eliminate this person, we are trying to automate the system using two models: Facial Expression Recognition and IoT devices model
IoT device Model:
We are planning to take the critical values using the IoT device, which is connected to the patient in real-time. Develop an ML model over it. Using anomaly/ outlier detection, we will find the unstable or not normal state of the patient to the doctor. Microsoft Azure is used as our cloud component. It is selected due to its rich support to the Health domain. For the notification part, we will notify users using two components. First is Power BI- where we will create a visualization of the vitals so that the doctors and other backend people can understand the cause of the unstable behaviour of the patient, second is the application notification- a mobile application to notify the doctor about the instability.
Facial Expression Recognition:
For this part, we have leveraged the readily available code. This module is developed using Keras and has the ‘pain’ attribute in it. With pain, it also detects these emotions: anger, disgust, fear, happiness, neutral, sad and surprise. But as a matter of fact, we are interested in ‘pain’. The camera will be set up to detect the patient expressions, which will be modelled over the cloud. Whenever out model outputs pain as the expression, we will notify the allocated person saying that the patient is in pain and needs immediate attention.
Bots are the future of engagement between the user and brand or let us just say, the fan and the celebrity!! We are planning to develop a chatbot for client eCommerce website using AWS Lex. We will chat with the user and try and find attributes like their age, gender and if they are interested in the respective product etc. Using these attributes to the apriori algorithm, we will get the products that are most suitable for the user. This will be done using the support and confidence of the algorithm. The bot will then offer these products to the user and help them directly to the website of the respective product. This will increase the engagement of the user with the website.
I had never worked in the USA before Ismile Technologies. The stand-ups are helpful as they give an idea of how we are doing in comparison to others, not in a demeaning way but as a healthy comparison.
They make sure that everyone is present on the On-Cam session right after the stand-up, this gives me a sense of accountability.
Hats off to Arif, for managing the entire team and making sure that there are no blockers on our way.
Not to forget my fellow interns and my teammates. I would not have enjoyed it as much as I did.