Close

Arunabh Singh

Data Scientist

Download Resume

About Me

An accomplished professional with more than four years of prolific experience in Data Science execution and delivery management pertaining to Financial Products and Digital Marketing. Well versed at establishing and integrating new internal or external datasets required to complete the analysis and enhance the validity, expectedness, and accuracy of models. Hands-on experience in comprehending data and reviewing results using statistical techniques, and preparing on-going reports accordingly. Deft in the development and implementation of databases, data collection systems, data analytics, and other strategies which facilitate competent data and reporting solutions. Well equipped with Python, AWS, Tableau, SQL, R, Keras,PyTorch and Excel

Experience

Aiztech Inc.

Sr. Data Scientist

  • Used various top-of-the-line ML models to accurately classify user personality data such as XG-Boost, Random Forest, SVM, keras based Neural Networks and other classifers.
  • Use user text data to understand the interests of Users using spacy and IBM NLU API.
  • Devised unique solutions to difficult data separation problems using various ML and statistical techniques and operational research.
  • Used IBM Watson Natural Language Understanding APIs to derive various user insights on users.
  • Used Flask to deploy end-to-end ML pipeline as an API service to be integrated in one of Aiztech’s core products.

Netcore Solutions Pvt. Ltd

Sr. Data Scientist

  • Ascertained the clients’ business requirements with regards to digital marketing
  • Prepared instructive and easy to understand proof of concepts on sample data for clients
  • Created 360 degree views for clients using MySQL to provide analytics solutions
  • Assisted the clients to reduce digital marketing costs by over 30% by providing the analyzed data for various micro-segments

Red Sky Technologies Pvt Ltd

Data Analyst

  • Implemented the unsupervised clustering of retail products using KMeans algorithm and featured engineering to sub-categorization of retail products
  • Wrote an extensive cataloging algorithm to match generic barcodes with in-store versions of those bar-codes that reduced manual entry costs by 20%
  • Wrote python scripts that used for specific regular expressions to extract specific invoice info as needed (This reduced manual checking and entry costs of invoices by 30%)

Pivot Management Consulting

Part Time Data Analyst

  • Adhoc analysis or Reporting as and when needed.
  • Worked part time to assist with project scope and efficiently interacted with various stakeholders on update and progress of some of the projects Pivot was handling.
  • Used SQL, Excel and other tools to analyze data and clean it and format it correctly.
  • Used Linear Regression to analyze and predict certain trends in data for various Clients.
  • Used PCA and K-means to cluster data and visualize it for certain projects.

Credit Suisse Business Analytics Pvt. Ltd.

Business Data Analyst

  • Reduced the losses by 25% of all trade orders for hedge funds and provided accurate analysis reports before trades got cleared by clearing house
  • Administered 800 weekly client meetings and provided the summary of top clients requirements and suggested the solutions on macro level which saved cost of resolution by 20%

Education

NYU tandon School of Engineering

Sept 2013 - May 2015

Masters in Technology Management

The Masters program at NYU involves an understanding of:

K J Somaiya Institute of Engineering and IT

Sept 2007 - May 2011

Bachelor of Engineering in Information Technology

Projects

World Bank Data Visualization Dashboard

This is a Data Visualization project which uses plotly and flask web framework to deploy a dashboard on heroku.

View Dashboard

View Project

AirBnB Listing Analysis and Data Visualization

The project involves analyzing AirBnB listings data for Toronto to answer business questions and write a blog with relevant findings.

View Tableau Visualization

View Blog

Dog Breed Classification

The goal of this project is to detect the breed of a dog using Convolutional Neural Network (CNN). The pipeline implemented will detect if an image has a dog or a human and then make a prediction of the dog’s breed or the breed, which the human most resembles.

View Project Repository

Flower Image Classification using Transfer Learning

This project involves the implementation of an image classification application using a deep learning model on a dataset of images.

View Project

Certifications

  • Tableau Desktop Specialist
  • Udacity Data Science NanoDegree

    • Skills

    Get in Touch