Chinmay Purav
415-***-**** adiodk@r.postjobfree.com LinkedIn San Jose, CA 95126
SUMMARY
Diligent Data Scientist with a passion for taking a deep investigative dive into data to uncover innovative solutions and opportunities for growth across various domains.
EDUCATION
San Francisco State University May 2020
Master of Science in Embedded Electrical and Computer Systems Galvanize, SF March 2020
Data Science Immersive Bootcamp
University of Nottingham, UK July 2015
Master of Engineering in Electronic Engineering (Hons) Uxbridge College, UK July 2012
Higher National Diploma in Electrical and Electronic Engineering Related Coursework: Intro to Database Systems, Engineering Communications, Engineering Management. SKILLS
Programming: Python, SQL, Git, AWS [EC2, S3], Docker. Libraries: Pandas, NumPy, Matplotlib, Seaborn, Scikit-Learn, LibROSA, SciPy, TensorFlow. Technical: Machine Learning, Hypothesis testing, A/B testing, Natural Language Processing, Data Cleansing, Feature Engineering, Statistical Data Analysis.
Databases: PostgreSQL, MySQL.
PROFESSIONAL EXPERIENCE
Data Analyst, Neoage Services [Remote] August 2020 – Present
● Leveraging critical data insights to inform progressive solutions to streamline decision making.
● Partnering with a client to flesh out the parameters and goals of the data science problem in the business.
● Located new process improvement opportunity.
Research Intern, Bonne Sante Meditech LLP [Pune, India] August 2016 – April 2017
● Researched and studied electro-chemotherapy instrument to design a cost-efficient version that rivaled market competitors.
● Modified probes for electro-chemotherapy to optimize device utility and attainment of tumor in difficult places.
● Created quarterly reports regarding the effectiveness of the electro-chemotherapy for different age group and genders.
● Regularly updated about 150 patients’ database to track patient progress, thus allowing lean organization to provide consistent primer care.
DATA SCIENCE PROJECTS
Are you Noisy Enough?
● The goal of the project was to classify audio files before cleaning the noisy audio files.
● Acquired dataset consisted of 400 audio files for each of 54 unique speakers and performed classification to label clean or noisy audio.
● Utilized LibROSA library to extract the features of audio files.
● Deployed Logistic Regression and Random Forest to achieve accuracy above 95%. Patterns in Independent Medical Reviews (IMRs)
● Performed hypothesis tests to examine the decisions overturned based on different types of diseases and age groups.
● Analysis of data consisting of 28,000 cases obtained from HealthData.gov concluded that denial of medical claim being overturned reduces with age.