Data Scientist Science

Karan Kishor Vyas

Boston, MA

Contact: 857-***-**** Email: ad5cxm@r.postjobfree.com LinkedIn GitHub

Experienced Data Scientist with 3+ years of work experience, proficient in Python, Data Science, and NLP, recognized for strategic leadership in leveraging Machine Learning and Deep Learning techniques for business impact. WORK EXPERIENCE

BYJU’S Pune, India

Sr BDA – Data Science July 2020 – Aug 2022

• Executed logistic regression, gradient boosting algorithms, and deep learning techniques to forecast user engagement on the Byju’s educational app, leading to annual cost savings of around $28,000 for the team

• Harnessed innovative NLP techniques, including sentiment analysis on product reviews, chatbot dialogue state tracking, and customer service automation, resulting in a 35% improvement in response times and efficiency

• Collaborated with Sales, IT, and Marketing to incorporate Python alongside scikit-learn for ML and data visualization libraries, leading to workflow optimization, improved system integration, and a 20% boost in productivity Decartem Pune, India

Associate Data Scientist May 2019 – June 2020

• Offered brainstorm POCs to simplify data integration processes and standard data feeds across systems, diminishing data latency from 8 days to 1 day adopting ETL tools

• Increased online order volumes by 200% with Python in designing 25+ A/B tests, automating data collection and applying statistical analysis and visualization techniques to refine key aspects of digital platform

• Analyzed data from 2.4 million monthly active app users, harnessing insights with libraries Matplotlib, Seaborn, and Plotly to create informative visualizations guiding marketing and product strategies, resulting in a 2x of average app engagement time, and a 3x rise in social media shares SKILLS

• Languages: Python, JavaScript, NoSQL, R

• Frameworks: NumPy, Pandas, Django, SciPy, Matplotlib, Seaborn, Keras, Scikit-Learn, NLP, Neural Networks, Deep Learning, Computer Vision, Data Mining, RDBMS, SQL, RESTful API, Statistical Modelling, Statistics, Data Visualization, TensorFlow, Tkinter, PyQt, PyTorch, Hugging Face, LangChain, LLM

• Tools: Power BI, Tableau, SQL, MySQL, MongoDB, GitHub, GitLab, Jira, MS Office

• Platforms: Linux, Windows, Web, AWS, GCP Cloud

ACADEMIC PROJECT EXPERIENCE

• Advanced Driver Assistance System (ADAS) Jan 2023 – May 2023

• Developed a real-time algorithm estimating Eye Aspect Ratio (EAR) in each frame leveraging TensorFlow, OpenCV, and PyQt reducing drowsiness alert response time by 40% and decreasing near-miss incidents by 35% enhancing driver safety

• Utilized Python, NumPy and Pandas, along with statistical tools, to analyze and mitigate driver drowsiness by monitoring eye blinking through Eye Aspect Ratio calculations, achieving 80.3% accuracy through rigorous steps including EDA, Feature Engineering and Hyperparameter tuning

• Leveraged facial landmark detection models trained on "300-W" to enhance robustness against varied illumination, head orientation, and facial expressions, leading to a 20% increase in detection reliability

• Harmoni Feel: The Emotive Music Matchmaker Aug 2023 – Dec 2023

• Implemented Emotion-based Music Recommender, employing Flask, advanced deep learning neural network architectures like BERT and GPT, and NLP techniques for data preprocessing, resulting in a 20% increase in user engagement through personalized music recommendations

• Engineered a user-friendly interface allowing users to input emotions or describe their day, processed through the emotion detection model, achieving a 25% reduction in response time for music recommendations

• Employed statistical analysis techniques such as hypothesis testing and regression analysis to evaluate user feedback and iteratively improve the recommendation algorithm, resulting in a 25% increase in recommendation relevance over time RESEARCH PROJECT

Research Paper Jan 2020

Vyas, Mr Karan, et al, 2020, in their paper titled "Driver Drowsiness Detection System Using Computer Vision" presented at the IRJET International Conference, introduced an innovative method employing recent advancements in facial landmark detection technology and statistical frameworks to detect driver drowsiness.

EDUCATION

Boston University Boston, MA

Master of Science - Artificial Intelligence Sep 2022 - Jan 2024 Relevant Coursework - Machine Learning (ML), Artificial Intelligence (AI), Data Science Tools (DS), Image Video Computing, Natural Language Processing (NLP), Big Data, Deep Learning (DL), Data Analytics for Machine Learning Sinhgad College Of Engineering Pune, India

Bachelor of Technology - Computer Engineering; Top 1 percentile Jul 2016 – May 2020 LEADERSHIP & ACTIVITIES

• Promoted within 8 months of joining in March 2021 for exceeding performance expectations, driving a 30% elevate in monthly customer acquisition, and achieving a 25% reduction in customer support response time at BYJU’S

• Exemplified strategic leadership in guiding a 60-member team as Mindspark’18 Secretary Head, ensuring event success through effective coordination and communication

Contact this candidate