S Vishal | CH18B020

Indian Institute of Technology Madras

Github: mew-two-github Codechef: vishal_padips Hackerrank: havocrulez
EDUCATION
Program Institution %/CGPA Year of completion
B.Tech Chemical Engineering IIT Madras, Chennai 9.52/10 2022(expected)
XII CBSE Chettinad Vidyashram 96.4% 2018
X CBSE P.S. Senior Secondary School 10/10 2016
SCHOLASTIC ACHIEVEMENTS
 Was part of 35 students selected from all over India for the final round of the Aditya Birla Scholarship, 2018.
 Secured an AIR of 1717 in JEE Advanced with a 99.14%ile and the Branch Opener of Chemical Engineering, 2018
 Secured an AIR of 709 in JEE Main-2018 among over 10 lakh participants with a percentile score of 99.99%.
 Was awarded the KVPY-SX, 2018 Fellowship conducted by IISc, Bangalore (secured an AIR 1063).
 CBSE merit certificate for securing top 0.1%ile in the country in Physics, Chemistry and Mathematics in AISSCE-18
RELEVANT COURSE WORK AND TECHNICAL SKILLS
 Programming Languages Known: C++, Python (and libraries in python: numpy, pandas, scikitlearn, keras), MATLAB
 Earned The Microsoft Technology Associate certification for Programming with Python; Know basics of git.
 Implemented Data structures and Algorithms in C++ in the Summer Project Workshop-2019(SPW) at IITM.
Mathematical Foundations of Data Science Machine Learning by Andrew Ng* Principles of Economics
Probability, Statistics and Stochastic Processes Computational Techniques Computational Techniques lab
WORK EXPERIENCE
Analyst Intern at AstraZeneca (May ’20 –July‘20)
 Worked on de Novo design of drug molecules to accelerate drug discovery for the therapeutic area of Oncology.
 Implemented an Actor-Critic model (a deep-RL framework) using RNNs [1] which optimizes potency of a molecule.
 Built and deployed a hyper-parameter tuned XGBoost regressor which predicts the pIC50[2] of a drug targeting AKT[3]
 Achieved an R2 value = 0.73 for regressor. Several potent drug candidates were generated by the final model.
PROJECTS
1. Complaints Classifier using AWS(DBS p-shift hackathon round-2) : (July-September 2019)
 An application which classifies voice based complaints for banks, converts it into text and summarises it.
 Deployed AWS tools: comprehend, transcribe and s3 bucket using the boto-SDK and built the app using python.
 Implemented the unsupervised ‘Text Rank’ algorithm for summarization using the nltk library.
2. CoVID-19 Prediction(Flipr Hackathon): (March –April 2020)
 Built a Machine Learning model to predict CoVID infection probability of a person in India using the given data.
 Applied One-Hot-Encoding of categorical parameters, filled empty data and did Time-series prediction of a label.
 Evaluated models on the mse [2] and concluded that a hyper parameter tuned Random Forest regressor is the best.
3. Mathematical Foundations of Data Science Course Project: (April 2020)
 Image Classification: Used SVD to classify 150 images into 15 classes and achieved an accuracy of 93.33%.
 Derived various insights, including the efficiency of the current lockdown from the “CoVID-19 in India” dataset.
4. Digit Recognizer(Kaggle beginner competition): (August 2020)
 Built a model for the classification of images of handwritten digits available in the popular MNIST dataset.
 Used an autoencoder for feature compression and a neural network for classification to achieve 97.5% accuracy.
5. SPW 2019 (Summer Project Workshop 2019, IITM): (May 2019)
 Built an E-Commerce application using heaps and hash tables in C++ to implement faster search for products.
POSITIONS OF RESPONSIBILITY
 Coding and Logic events coordinator in Shaastra 2020 (April ’19 – January ‘20)
 Ideated and executed 2 brand new events: Debug the Code & E-Contest; witnessed a total footfall of 300 teams.
 Set up problems, tests and judges. Responsible for completion of MoUs with Wells Fargo and American Express.
EXTRACURRICULAR ACTIVITIESS
 Enthusiastic participant in competitive coding; Gold-Badge (6 stars) in Hackerrank Problem Solving.
 S-NET [4]: Organized seminars, cleanup events and awareness drives to promote sustainability within the institute.
 Mentoring 8 freshmen as a student mentor and helping them get involved in co-curricular activities in the institute.
*
- MOOC [1]
–Recurrent Neural Networks [2]
- Protein Kinease B
[3] [4]
- Mean squared Error -Sustainability network