Data Scientist with ML and Deep Learning experience

EDUCATION AND HONORS

Indian Institute of Technology Roorkee, Uttarakhand, India

Master of Technology in Electronics and Communication Engineering (M.Tech)

July 2023

Major: Communication, Network and Signal Processing
Study in India (SII) Scholarship
Relevent Coursework: Machine Learning, Linear Algebra, Compressed Sensing, Information Theory and Coding, Detection and Estimation Theory

M. S. Ramaiah Institute of Technology, Bangalore, India

Bachelor in Engineering Electronics and Communication (B.E)

Aug 2017

EdCIL (Indian Government) Scholarship
Relevent Coursework: Neural Networks and Fuzzy systems, OOP, Data Structures, Probability & Discrete Mathematics, Data Science Statistics

Work Experience

Faculty - Python, Artificial Intelligence, and Data Structure & Algorithm | BCS IT Program, KFA Business School & IT, Kathmandu

11/2023 - Present

Teaching critical computer science subjects including Python (PHN127), Artificial Intelligence (AFI353), and Data Structure & Algorithm (DSA-233) to 2nd, 5th, and 6th semester BCS IT students. Leading engaging lectures, hands-on labs, and project-based assessments aimed at integrating foundational knowledge with practical application in AI, and advanced programming techniques.

Senior Officer IT, KFA Pvt. Ltd., Kathmandu

03/2018 - 07/2021

Applied analytical methodologies to derive actionable intelligence, enhancing the organization’s ability to make informed and data-driven decisions.
Spearheaded strategic IT initiatives, providing foresight and senior-level guidance to align technology efforts with organizational goals.
Led the design of robust technical architectures, ensuring alignment with organizational objectives.
Represented the IT department in diverse projects, guaranteeing adherence to architectural requirements and principles.
Integrated data analysis into decision-making processes, leveraging insights for strategic planning.

Intern, LI2 Innovation, Bangalore

01/2016 - 08/2016

Designed and developed various embedded modules and projects, including a prototype for an “Obstruction Detecting robot with Smart sidelights” using Raspberry Pi and Arduino platforms.
Prepared a project on Arduino-based robotics titled “Face detection robot.”
Built an Arduino-based “Solar-powered Motion Sensor Smart Night Light.”
Built a prototype for web-page-based Home Automation using Raspberry Pi.
Organized, coordinated, and conducted training and workshops on electronic modules at education and entrepreneur levels.

Projects

Customer Churn Prediction using MLOps

9/2024 - 10/2024

Developed a machine learning model to predict customer churn in an online retail context, utilizing a diverse range of features such as user demographics, browsing behavior, transaction history, and membership details. Implemented an ETL pipeline to extract, transform, and load data from MongoDB Atlas, and applied data preprocessing techniques such as missing value imputation, feature engineering, and data transformation to improve model accuracy.
Explored multiple machine learning models, with a focus on LightGBM, to predict customer churn risk scores ranging from 1 to 5. Leveraged techniques such as feature scaling, hyperparameter tuning, and cross-validation to optimize performance, providing the business with a robust solution for predicting and mitigating customer churn.
Link to access the GitHub Repository and DagsHub Repository

Application Tracking System Web App

12/2023

Created an ATS web application to help in the screening of resumes efficiently. Combined the recently developed Google Gemini AI models with OpenAI to deliver its analysis and suggestions in regards to job descriptions. Included convenient options like the possibility to upload the resume in the PDF format and the option to input the job description in the text.
Applied superior NLP methods for extracting and analyzing the quintessential resume data to the existing application. Used custom CSS for its proper and appealing look on user interface. It also provides an option of general feedback that will enable the users to have an insight of the resume match percentage, missing keywords, and profile summary that is very essential to the job seekers.
Link to access the web application Application Tracking System Web App

Non-Invasive Neuroimaging-based Automated Epilepsy Detection

01/2022 - 06/2023

Employed innovative N-cylinder-based feature engineering techniques combined with advanced machine learning methodologies. This project significantly advanced epilepsy detection using non-invasive neuroimaging data, facilitating early diagnosis and refining treatment strategies.

Riemannian Approach-based Depression Classification using Transfer Learning for MEG Signals

04/2022 - 06/2022

An award-winning project for the BIOMAG 2022 ‘Ketamine for Depression’ challenge developed a novel Riemannian approach-based classification framework for accurately identifying depression using Transfer Learning techniques applied to MEG signals.
By leveraging the power of Transfer Learning and incorporating Riemannian geometry, our framework effectively captured the unique characteristics of depression in MEG signals, surpassing existing methodologies and holding promising implications for improving diagnosis and understanding of depression.

Text-Based Classification for Document Categorization

02/2021 - 03/2021

Developed a machine learning NLP model for document categorization, emphasizing the use of textual data. The project involved text preprocessing, feature extraction using TF-IDF and Bag of Words, and the implementation of classification algorithms such as SVM, Random Forest, Naive Bayes, and xgBoost.
Explored various classification algorithms, with SVM exhibiting superior performance. Identified challenges in distinguishing classes with overlapping vocabulary, emphasizing the need for nuanced feature extraction techniques for more accurate categorization.

Automated Data Management for Quest Pharmaceuticals, Nepal

07/2021 - 08/2021

Pioneered the development and implementation of a user-friendly and automated Excel form using VBA during my tenure at KFA. Transformed data management for Quest Pharmaceuticals, Nepal, contributing to enhanced efficiency and streamlined processes.

Google Analytics Analysis for Online Courses

08/2020 - 09/2020

Analyzed user behavior within KFA’s online courses using Google Analytics data. Investigated various parameters such as course details, website interactions, and engagement patterns. The study aimed to uncover valuable insights into user preferences and interactions to enhance the overall online learning experience.

Secure Online Shopping using Fingerprint Verification

08/2016 - 05/2017

Developed a secure online shopping platform leveraging Python and the Minutiae Matching Algorithm (MMA) for real-time fingerprint-based authentication. Replaced OTP-based verification to reduce dependency on mobile devices, enhancing security against unauthorized transactions. Built a full-stack application with MySQL for secure data storage, XAMPP for server management, and a user-friendly web interface using PHP, HTML, and CSS.
Link to access the GitHub Repository

Skills

Technical Skills

Predictive Modeling (Experienced), Data Modeling, Unit Testing, Natural Language Processing, Critical Analysis, Statistical Analysis, Predictive Analysis, Machine Learning Algorithms, Computer Vision.

Mathematical Foundations

Statistics, Linear Algebra, Calculus, Probability

Tools and Software

Python, C, Matlab, Scikit-Learn, Numpy, TensorFlow, SciPy, OpenCV, Streamlit, Flask, Git and Github, Docker, MySQL

Programming Skills

Object Oriented Programming (OOP), Data Structures and Algorithms (DSA)

Papers & Publications

S. K. Varun, T. K. Reddy Bollu, and S. Shrestha, “Automatic Seizure Detection Employing Machine Learning-Based Late Fusion Techniques Over Behind-the-Ear and the Scalp EEG Signals,” 2023 IEEE 4th Annual Flagship India Council International Subsections Conference (INDISCON), Mysore, India, 2023, pp. 1-4, doi: 10.1109/INDISCON58499.2023.10270088 link.
Acknowledged contributor in the paper titled “A Meditation Based Cognitive Therapy (HMBCT) for Primary Insomnia: A Treatment Feasibility Pilot Study,” published in Applied Psychophysiology and Biofeedback (2023). The acknowledgment highlights my instrumental role in statistical and mathematical analysis, contributing crucial insights to the research methodology. This acknowledgment underscores my proficiency in applying advanced statistical techniques to enhance the study’s robustness. [DOI: 10.1007/s10484-023-09586-2] link

Achievements and Recognitions

BIOMAG Data Analysis Competition 2022

Achieved 1st position in the Ketamine in Depression Machine Learning Challenge, an international competition. Excelled in the classification problem, showcasing expertise in machine learning and a strategic approach to problem-solving link.