Resume

Sign in

Data Analyst Engineer

Location:
Milpitas, CA
Posted:
May 17, 2020

Contact this candidate

Resume:

AKANKSHA NARWANI

+1-405-***-**** adc9ut@r.postjobfree.com linkedin.com/in/akanksha-narwani

EDUCATION

Master of Science - Management Information Systems with Data Science (GPA-4.0) May 2020 Oklahoma State University, Stillwater, OK

• Received Academic Excellence Recognition by the Honor Society of Phi Kappa Phi

• Received Ramesh and Usha Sharda Family Endowed Scholarship Bachelor of Engineering with Honors - Electronics and Telecommunications (GPA-3.8) May 2017 Chhattisgarh Swami Vivekananda Technical University, Bhilai, India PROFESSIONAL EXPERIENCE

Tesla, Fremont, CA - Data Engineer Intern January 2020 - May 2020

• Built a custom real-time data streaming and analytical dashboarding application for over 350 Tesla service-centers. Used React for designing dashboards, Node.js for creating secured and custom API endpoints, and Splunk for analyzing log files.

• Engineered a re-usable and custom ETL solution using Airflow (open source workflow manager) to save up to 80% of Data Engineer’s time in writing code.

• Spearheaded development of chat-bot using open source conversational AI tool (Rasa) to provide access to requested data and hence manage and resolve 85% of end-user interactions. American Fidelity Assurance, Oklahoma City, OK - Data Analyst Intern May 2019 - August 2019

• Proposed a model to accomplish 40% decrease in claims-fraud rate: Performed predictive analysis and anomaly detection to identify fraudulent behavior in claims using Decision Trees and Naïve Bayesian Classifier.

• Shortened data retrieval time by 85% using dynamic API accounts integrated with Parallel Processing in the data pipeline.

• Employed statistical modeling techniques to study the impact of variables like claim history in the neighborhood, weather patterns etc. on premiums for policyholders.

Oklahoma State University, Stillwater, OK - Research Assistant January 2019 - Current

• Realized 30% increase in the accuracy over baseline by developing statistical models (Support Vector Machine and Decision Tree Classifier) to study sentiments of US citizens on proposed changes in national laws.

• Created Python scripts to scrape and parse unstructured text data from Twitter using BeautifulSoup, RegEx and Tweepy packages and devised text mining on over 500,000 tweets about administration for Sentiment Analysis. Oklahoma State University, Stillwater, OK - Data Analyst August 2018 - December 2018

• Scraped and modeled university data from various websites using Python’s BeautifulSoup library for providing it to various ranking publications. The world ranking of Oklahoma State University in the publications went up by over 80 positions.

• Employed Regression Trees, Lasso and Ridge Regressions to detect the key factors contributing to University Ranking. Innolat Technologies, Raipur, India - Software Developer Intern January 2017 - January 2018

• Achieved an 8% revenue growth in a quarter by a 20% increase in student enrollment by establishing interactive KPI visualization dashboards to generate marketing signals of targeted customers.

• Computed scoring algorithms and designed the dashboard of ‘Score Board’ module for the website using AngularJS and CSS. TECHNICAL SKILLSET

Programming: SQL (Advanced), Python (Advanced), R, JavaScript, VBA, Unix Shell, Node.js Machine Learning Algorithms: Decision Tree, Regression and Classification, Neural Network, Ensemble B&B, Kernel Methods & SVMs, Markov Decision Process, Game Theory, Bayesian Learning Statistical Techniques: Regression Analysis, Time Series Forecasting, Monte Carlo Simulation, Hypothesis Testing, A/B Testing Applications & Databases: Tableau, Power BI, Qlik, Oracle SQL, MySQL, PostgreSQL, Apache Spark, AWS Redshift, Splunk, Vertica ACADEMIC PROJECTS

Flower Image Classifier using Deep Learning August 2019 - September 2019 Built a Neural Network using TensorFlow on 32,000 images to classify 102 flower species with less than 0.1% error. Developed an application using Python Shell to input images and output the predicted class of the flower. Identifying Customer Segments for a Mail-Order Sales Company May 2019 - July 2019 Accomplished 30% increase in customer response rate: Clustered the population features using MiniBatchKMeans algorithm and PCA to identify potential customers in the UK. Used Silhouette Coefficient to evaluate models.



Contact this candidate