Swastik Udupa
• swastikudupa@[Link] • +91 9740641023 • #A205, Mega Tower 3, NITK, Mangalore, India
Education [Link] in Information Technology, Final Year 2015 - 2019
National Institute of Technology Karnataka, Surathkal, India
• CGPA of 8.01/10 (May 2018)
Class 10 - Little Rock India School, Udupi (CBSE) - 10.0 CGPA 2012 - 2013
Class 12 - Poorna Prajna Pre University College, Udupi (KSEEB) - 95.6% 2014 - 2015
Experience Software Developer Intern, Hashlearn, Bangalore May-July, 2017
Performed integration testing for some of the major flows in the [Link] with django, postgresql,
redis and tools like Postman to complete performing the tests for the entire flow.
Designed and implemented machine learning algorithms such as KNN to provide solutions to tutor availability
and rating predictions. The new methods implemented were found to produce 50% more accurate results
than the methods originally being used.
Software Developer Intern, Amazon, Chennai May-July, 2018
Upgraded the metrics recording system in place to a newer platform and designed schema to log the required
data points. Changed the original theme used in the out of box experience screens in Fire tablets to a newer
theme. Created new pages, re-designed existing pages and designed reusable and flexible components usable
by any screen wanting to change the theme in the future. The newer theme was designed as per the redlines
document that was provided and involved taking several independent calls on design and implementation.
Work done had a great impact since millions of customers using the fire tablets will get to see the screens
using the newer theme.
Technical Languages - C, Python, Java, JSP, JS
Skills Tools/Frameworks - Git, Django, STL, MySQL, OpenMP, Flask, Keras, Android
Selected • Patient procedure sequence prediction 2018
Projects An ongoing project where RNN, LSTM models are applied on Mimic-3 Datataset to predict the procedure
codes by taking the didease sequence as input. The architecture of the model is designed is a such a way
that the model results in a good accuracy for procedure detection for a new patient with a disease.
• Malware Classification using Deep belief network 2018
Classification of malware and benign android apks by using deep belief network on the processed apk
data set. The model has been able to produces upto 99% accuracy on the available dataset.
• Healthify 2017
An android application which enables the user to view the nearest hospitals and send messages to contact
and nearest hospital giving exact location of the user in case of a medical emergency.
• Content Management System 2017
Designed a content management system application using Java which is used to store and manage employee
details. Java Swing library was used to build the GUI and could be connected to the database through
which the details where stored and retrieved.
• Command line IMDB Scraper 2016
Developed a Linux command line utility that can scrape [Link] and return result depending on the
user’s query. Incorporated a movie guessing game with the same.
Achievements • Selected for National Talent Search Examination Scholarship in 2010.
• AIR 4999 JEE Mains in 2015 and rank 244 in Karnataka Common Entrance Test.
Extra • Executive member of ACM, NITK Chapter.
Curricular
• Executive member of LSD(Literary, Stage and Debating Society).
Activities
• National runners up of TCS Rural IT Quiz, 2012.
• First in regional and third in state at TCS Tech Bytes, 2016.
