Data Engineer

Deepanshu Parihar

*******.*@*****.***.*** Ph: 617-***-****

https://devpost.com/DeepanshuParihar Boston, MA-02215 www.linkedin.com/in/pdeepanshu/ Available from: May 2020 Education

Northeastern University, Boston, MA, USA September 2018–Present Khoury College of Computer Sciences Expected graduation: December 2020 Candidate for a Master of Science in Data Science

• Related Coursework: Data Management & processing (R), Algorithms, Supervised Machine Learning, Robotics Science and Systems, Natural Language Processing, Special Topics in AI: Imaging & Deep Learning

SRM University, India August 2012–May 2016

Bachelor of Engineering in Electronics & Communication

• Related Coursework: Probability & Random Processes, Linear Algebra & Statistics, Discrete Mathematics. Technical Knowledge

Languages: Python, R, Java

Technologies: Python Stack (Jupyter, PySpark, PyTorch, Flask, PyCharm, Pandas, NumPy, Sci-kit Learn, OpenCV, Gensim, NLTK, Matplotlib, Plotly).

Databases (Relational: MySQL, NoSQL: MongoDB), Tableau, RStudio, Eclipse, Hue, Google Colab, Google Cloud Platform, Linux (Ubuntu), Windows. Work Experience

Stanley Black & Decker, New Britain, Connecticut, USA June 2019 – December 2019 Data Scientist Coop

• Built data pipelines in pyspark on an AWS cloud to transform and feed ERP data from hundreds of manufacturing plants to dashboards that provide up to date analytics to procurement managers to negotiate better deals with vendors based on fluctuations in the global commodity and forex indices.

• Created a model for classifying transactions made by the global supply chain team of millions of values with vendors based on company’s four level taxonomy.

Teaching assistant - Northeastern University, Boston, MA Jan 2020-Present

• Khoury College of Computer Sciences - CS 6220 Data Mining Techniques Cognizant Technologies Solutions, Chennai, India June 2016 - July 2018 Programmer Analyst / Product Engineer

• Achieved data reduction of 100 GB by extracting unique frames in a video using FFMPEG.

• Built an object detection and tracking prototype by leveraging OpenCV in Python for a client. The prototype was part of big presentation which was successful in getting the project. Academic Projects

Santander Customer Transaction Prediction January 2019-April 2019

• Built classification model using Light GBM to predict customer transaction resulting in 0.92 ROC-AUC.

• Implemented Neural Ordinary Differential Equation research paper using PyTorch for a more accurate model.

• Used feature engineering on 200 variables based on bagging to obtain 30 most relevant features. Analysis of NHIS data and building associated software components Oct 2018-December 2018

• Analysis of NHIS data to understand sleep quality, affordability of medical equipments of people in USA.

• Built a chatbot to obtain normal statistics of the data and other relationships.

• Developed an R Package to export different cuts of analysed data and necessary functions used by chat bot.

• Built a classification model to detect affordability of components of healthcare. Hackathons

• Won the Colgate Data Science Challenge of estimating product sale price based on location, brand and ingredient of the product at Hack Rutgers in Fall 2019.

• Won “Best Constellation Brands Consumer Experience” and a special mention for Best Brick Hack category for building an app to scan QR codes and give AR representation of the product at Brick Hack 5 in Spring 2019.

Contact this candidate