Omkar Mane

About

Learn more about me

Machine Learning Engineer

Passionate ML Engineer with expertise in developing and deploying advanced machine learning models.

Birthday: 3 May 2001
Phone: +571 245 7890
City: Washington DC, USA

Age: 23
Degree: Master of Science
Email: omkarbalasaheb.mane@gwu.edu

With a solid foundation in computer science and hands-on experience in machine learning, I am dedicated to creating innovative solutions that drive business success. My background includes developing sentiment analysis models, leading real-time gameplay assistance projects, and enhancing voice assistant systems. I thrive in fast-paced environments and am always ready to tackle new challenges. My proficiency in Python, TensorFlow, and data analysis enables me to transform complex data into actionable insights. Let's connect to explore how we can drive technological advancements together.

Skills

Python 100%

TensorFlow 90%

Data Analysis 92%

Docker 85%

SQL 90%

AWS 80%

Interests

Cooking

Tech Blogs

Fitness

Volunteering

Resume

Check My Resume

Sumary

Innovative and detail-oriented Machine Learning Specialist with hands-on experience in developing and deploying advanced machine learning models. Skilled in enhancing data quality and creating efficient solutions for sentiment analysis and voice assistance. Adept at collaborating with cross-functional teams to deliver impactful AI solutions from concept to implementation.

2020 F St NW, Washington, DC
(571) 245 3562
omkarbalasaheb.mane@gwu.edu

Education

Master of Science in Computer Science

2023 - 2025

The George Washington University, Washington, DC

Coursework includes Machine Learning, Large Language Models, Neural Networks, and Deep Learning.

Bachelor of Engineering in Electronics & Tele-Communications

2018 - 2022

Savitribai Phule Pune University, Pune, IND

Specialized in electronics with a focus on MATLAB, Control Systems, Digital Signal Processing, Digital Analysis, Embedded Systems, Artificial Intelligence, Machine Learning, and SQL.

Professional Experience

ML Intern

APR 2023 - AUG 2023

Way Down South, Pune, IND

Conducted comprehensive sentiment analysis through data fetching and exploratory data analysis leveraging Pandas and NumPy frameworks, uncovering valuable insights to inform decision-making.
Developed a sentiment analysis model using TF-IDF vectorization and SVM, improving accuracy by 15% with 92% precision and 88% recall. Enhanced customer feedback analysis and informed data-driven decisions.
Led testing and validation processes, accomplishing a 20% increase in model reliability. Sentiment analysis model exceeded industry performance benchmarks, ensuring high-quality insights for decision-making.

Programmer Analyst

JAN 2022 - APR 2023

Cognizant, Pune, IND

Directed application of diverse data transformations, collaborating with a team to improve data quality by approximately 30% through thorough cleansing, aggregation, and structuring processes.
Coordinated within a team to leverage Business Objects Data Services functions such as lookup, substring, and to_date, optimizing data retrieval and formatting to achieve a 25% efficiency boost.
Created and applied advanced data validation checks in data services, attaining a 95% validation success rate and mitigating potential inconsistencies and errors in downstream systems.

Projects

My Projects

Enhanced Voice Assistant

This project involved developing and executing a comprehensive testing protocol for a Python-based voice assistant, leading to a 20% improvement in speech recognition accuracy and reducing task execution response time to under 1 second. By collaborating with a team to integrate advanced natural language processing capabilities, the assistant facilitated more complex user interactions and increased user satisfaction by 15%. The addition of new features, such as real-time weather updates and music streaming, contributed to a 30% boost in user engagement.

Sentiment Analysis

This project focused on conducting a comprehensive sentiment analysis to derive actionable insights from customer feedback. Utilizing Pandas and NumPy for data fetching and exploratory analysis, the project employed TF-IDF vectorization and SVM to develop a sentiment analysis model, achieving a 15% improvement in accuracy with 92% precision and 88% recall. The model's reliability was enhanced by 20% through rigorous testing and validation, surpassing industry benchmarks and providing high-quality insights to inform strategic decision-making.

Valorant Mentor using CNN

This project involved the development and deployment of a CNN-based machine learning model to provide real-time gameplay assistance. The system was enhanced through effective data acquisition and preprocessing, including color/image augmentation and duplicate removal, which improved processing speed by 30% and increased model accuracy by 12%. The model was integrated into a user-friendly application using Flask and React.js, leveraging TensorFlow and OpenCV for real-time object detection. Continuous monitoring and model retraining were implemented to ensure sustained accuracy and performance, achieved through collaborative teamwork.

Chronic Heart Failure Prediction

This project aimed to develop a machine learning model for predicting chronic heart failure (CHF) based on clinical data. Using the UC Irvine Machine Learning Repository dataset, the project involved extensive data preprocessing, including feature selection and class balancing through Random OverSampling. A Naive Bayes classifier was employed and tuned, achieving a 73% accuracy with moderate performance in predicting CHF-positive cases. The model's insights and applications highlight its potential as a decision support tool for early intervention and personalized treatment. Recommendations include exploring advanced techniques and incorporating domain expertise to improve predictive accuracy and clinical utility.

Text-Based Emotion Recognition

This project focused on developing machine learning models for emotion recognition in text, employing Support Vector Machines (SVM) for classification and K-Means clustering for pattern discovery. Using the EmotionLines dataset, the project involved preprocessing text data and transforming it with TF-IDF. The SVM model was optimized and evaluated, achieving moderate performance with a 50% accuracy rate, while K-Means clustering provided insights into emotional patterns with a silhouette score of 0.051. The findings highlight the strengths and limitations of each approach, with recommendations for combining methods and enhancing model performance.

Lip Reading with Deep Learning

This project focuses on developing a deep learning model capable of reading lips using Python and TensorFlow. The model leverages a CNN-RNN architecture to process video frames of lip movements and accurately predict spoken words. By training on large datasets of spoken sentences, the system aims to enhance the accuracy and reliability of lip-reading technology, offering potential applications in communication aids and security systems.

Contact

Contact Me

My Address

2020 F St NW, Washington, DC

Social Profiles

Email Me

omkarbalasaheb.mane@gwu.edu

Call Me

+1 (571) 245 3562