Machine Learning Data Science

MANASEE RAMCHANDRA PARULEKAR

+1-213-***-**** ********@***.*** https://www.linkedin.com/in/manasee-parulekar/ https://github.com/Manasee27 EDUCATION

University of Southern California - Viterbi School of Engineering GPA- 3.5/4 Los Angeles, CA Master of Science - Electrical and Computer Engineering (Machine Learning & Data Science) May 2024

• Relevant courses: Machine Learning, Web Technologies, Probability, Optimization for Information and Data Sciences University of Mumbai- Vivekanand Education Society’s Institute of Technology GPA- 8.87/10.00 Mumbai, India Bachelor of Engineering – Electronics and Telecommunication Engineering May 2022 TECHNICAL SKILLS

Programming Languages: CSS, XML, Python, R, C++, Go, Java, MATLAB, JavaScript, C#.NET Software: BigQuery, Git, Linux, AWS, Jupyter Notebook, LTspice, Android Studio, Docker Deep Learning Frameworks: TensorFlow, PyTorch, Keras, MediaPipe, Transformers Web Frameworks: Angular, React, Node.JS, Agile, Flask, Django, Spark Tools and Databases: Snowflake, MySQL, PostgreSQL, Kafka, Hadoop, MongoDB, Power BI, Tableau

WORK EXPERIENCE

AI CAMP Palo Alto, CA

Machine Learning Applications Intern [website- https://faculty-finder.streamlit.app/ ] May 2023 – Nov 2023

• Developed a React web application to accurately extract faculty data from 2,000 high schools across the United States, successfully retrieving 100,000+ contacts using Selenium and Puppeteer

• Used Node.js for MongoDB connectivity and embedding generation. Configured GitHub Actions to ensure cloud execution

• Employed Pinecone with a 764-dimensional vector space using the Sentence Transformer for advanced embeddings. Conducted cosine similarity searches and leveraged OpenAI GPT-3.5 Turbo, LLMs with zero-shot prompting for refined results THE SPARKS FOUNDATION Mumbai, India

Data Science Intern Jul 2021- Sept 2021

• Designed a machine-learning model for stock price prediction utilizing textual and numerical analysis of data from Indian stock exchange platforms (BSE/NSE). Implemented sentiment analysis and ETL to obtain productivity of textual data.

• Compared and fine-tuned three models to minimize mean square error by 12%

• Implemented Python Flask server to serve model with HTTP requests. Created web application using HTML, JavaScript, and CSS VESIT Renaissance Cell, VESIT Mumbai, India

Software Engineer intern Dec 2019-Jan 2020

• Devised a web application from scratch to automate flight slot allocation using Bootstrap tech stack

• Organized backend tasks in Laravel framework and stored flight information in phpMyAdmin database

• Proposed and executed FLASK framework for deploying Machine Learning model in web server by achieving an accuracy score of 80% PROJECTS

Scalable YouTube Data Analytics pipeline using AWS Apr 2024

• Designed a scalable system to collect data from multiple sources, including YouTube's API, and provide visualizations on views, likes, comments, and other engagement data

• Developed a Data Lake on AWS S3, integrating Glue and Athena to enhance data processing speeds and reduce retrieval times by 30%

• Generated analytical dashboards in AWS ‘QuickSight’,’Airflow’ and implemented a serverless ETL architecture using AWS Lambda State-of-the-art aid for the visually impaired [GitHub - SAFE-STEPS ] Jul 2023

• Spearheaded a project employing YOLO v5 for object detection, optimizing with transfer learning and data distillation techniques, resulting in a 15% improvement in detection accuracy

• Scaled Intel's MIDAS model with traditional object detection algorithms, enabling precise 3D object localization from 2D depth maps

• Seamlessly deployed PyTorch optimized model through Hugging Face's inference API within a Flask-driven web interface Website and Android application to search for events using Ticketmaster APIs Mar 2023

• Accomplished a user interface feature allowing event marking as 'Favorites', with seamless routing to display a consolidated list of favored events. [ demonstration - https://youtu.be/Ap-oznnTxDo ]

• Implemented backend applications using JavaScript/Node.js on server side with Express framework

• Architected a responsive front-end leveraging Angular, TypeScript, AJAX, and Rest APIs for optimal client-side performance Sign Language (ASL) translation using a chrome extension for Google Meet [GitHub- ASL_Chrome_Extension ] Feb 2022

• Captured video frames from Google Meet are sent to MediaPipe framework to predict 50 unique hand gestures

• Adopted MongoDB to store hand sign coordinates in form of arrays and send it to Node JS server and then chrome extension. Designed an ML model with an accuracy more than 90%. Launched application on cloud using GCP

• Enhanced system by incorporating text-to-speech and leveraging Lang-8 with a deep learning model for grammatical accuracy PUBLICATIONS

• "Sign Language Translation Using a chrome Extension for Google Meet". Published and presented Jul 2022 paper at the 2022 IEEE Region 10 Symposium (TENSYMP) https://ieeexplore.ieee.org/document/9864398

• "Air Traffic Control Database Management System for Seamless Transit". International Research Journal Jan 2022 of Engineering and Technology, Volume 9 Issue 01

LEADERSHIP & COURSES

• AWS Certified Cloud Practitioner CLF-02, (UDEMY, Expected Completion: June 2024)

Contact this candidate