Vivekanand Education Society's Institute of Technology, Mumbai
Vivekanand Education Society's Institute of Technology, Mumbai
Dec 2021 - May 2025
B.E. in AI & Data Science, Honours in Cybersecurity
AI Courses: Artificial Intelligence, Maachine Learning, Deep Learning, Reinforcement Learning
Data Science Courses: Data Warehousing and Mining, Big Data Analytics, Social Media Analytics, Statistics
CS Courses: Data Structures, Algorithms, DBMS, OS, Web Development, Software Engineering
GPA: 9.6/10
Christ Academy, Navi Mumbai
July 2021
Completed HSC Class 12 with Distinction
Percentage: 97.67%
Christ Academy, Navi Mumbai
May 2019
Completed SSC Class 10 with Distinction
Percentage: 89%
Veermata Jijabai Technological Institute
Research Intern
Jan 2025 - April 2025
Contributed to an autonomous vulnerability patching system by developing AI agents that analyze code snippets and identify security vulnerabilities with a detailed analysis using a CVE-based knowledge base.
Fine-Tuned CodeT5 models for classifying vulnerable and safe code.
Reliance Industries Ltd.
ML/Data Science Intern
May 2024 - July 2024
Worked on creating a Bidirectional LSTM model for Gas Sales Forecasting.
Used Angular, OpenAI API, and Selenium to implement a RAG System for Dashboard Analysis & Interpretation.
Spartificial
ML Research Intern
May 2023 - July 2023
Worked on the research project titled: 'Detecting Volcanoes on Venus'.
Created a labeled dataset of volcanoes on Venus using satellite images from NASA.
Built a CNN-based computer vision model with transfer learning to detect volcanoes.
Report: Detecting Volcanoes on Venus.
Conference
[1] S. Oswal, R. Bhalerao, A. Kulkarni, “Optimizing Biomass Forecasting and Supply Chain: An Integrated Modelling Approach,” Advanced Computing, IACC 2023. Communications in Computer and Information Science, vol. 2053, Springer, Cham, pp. 57-71, 2024.
ISTE-VESIT
Chief Technical Officer
Jan. 2023 - April 2025
Speaker for a workshop on React, Flask, and a simple recommendationation system; attended by 100+ people.
Developed a full-stack web application: Spotify Clone, using Flask & React along with a recommendation system.
Speaker for a workshop on creating and publishing an AR game using Meta's Spark AR.
Organized a 24-hour hackathon with over 400 participants.
Developed and fine-tuned Small Language Models (SLMs) for Micro, Small, and Medium Enterprises (MSMEs), achieving competitive performance with lower resource requirements compared to larger models.
Implemented scripts for data preprocessing, model fine-tuning using HuggingFace transformers and PEFT, and evaluation of SLM performance.
Generated an augmented reasoning dataset from ECInstruct using quantized DeepSeek-R1 models to enhance model training and evaluation.
Techonologies Used: React, Node.js, Express, MongoDB, Unity
Developed an integrated system aimed at enhancing cognitive retraining for children with developmental disabilities.
Implemented features that provides retraining to the patients in the form of interactive games using Unity.
Developed systems for therapists to assist them in making more effective analysis by presenting summarized visual representations of patients activities.
Techonologies Used: React, Node.js, Express, MongoDB, Unity
Enabled: Connections Matter
Created an AI based community platform for Persons with Disabilities (PwD) with various AI based accessibility features, allowing for easy use for the PwDs.
Presented the idea and app at the regional conference finals as a part of eYantra eYIC competition held by IIT Bombay.
Techonologies Used: E-Yantra, AI, Flutter, Ideation
Made a full fledged template for a generic quick setup of Session Based Authentication.
Implemented protected routing with a clean UI interface with proper backend implementation of cookies and storing sessions.
Added support for HTTPS testing by enabling a flag and providing the appropriate certificates.
Techonologies Used: React, Node.js, Express, MongoDB
Predicting Forest Fires using AI/ML
Aug 2022 - Apr 2023
Explored and understood time series forecasting and analysis using variuos supervised ML algorithms.
Implemented models using various ML Algorithms such as XGBoost & fbProphet.
Built a model that could predict a forest fire with 99.2% accuracy within India over the course of 2022.
Techonologies Used: Scikit-learn, AI, ML, XGBoost
Tree Walk Interpreter
Implemented a basic tree walk interpreter in Java followed from Crafting Interpreters.
Learned about various design patterns used in compilers.
Can interpret basic control flow statements.
Techonologies Used: Java
Made an agent that learns how to play the snake game using Reinforcement Learning.
Used Deep Q Learning for the agent to learn how to play snake.
Implemented the game using PyGame which acts as the environment for the agent.
Techonologies Used: Python, PyTorch, Reinforcement Learning
Created a Simple Desktop Application that interacts with a database for auto filling the cost of drugs.
Implemented the option to print invoices.
Techonologies Used: Java, Swing, Apache, MySQL
Programming Languages
Frameworks
E-Yantra Innovation Challenge
April. 2023
Regional Conference Finalist
Shell.ai Hackathon for Sustainable and Affordable Energy
October. 2023
Ranked 12th in University Edition
IIIT Sricity National Healthcare Hackathon
December. 2023
1st Runner Up in IIIT Sricity National Healthcare Hackathon
Placement Co-ordinator
Organized and Assisted the Placement Cell with conducting On-Campus Placement Drives
VESIT Basketball Team
Participated in various Inter College Events as a part of the college basketball team