Ameya Karnad

About Me

Hi, my name is Ameya Karnad, and I am Software Engineer working at Google in Mountain View, California. I graduated from Columbia University with a Master’s Degree in Data Science in December 2019.

I am interested in the Applications of Data Science in the fields of computer science, technology, social sciences, public policy, political science and education. Recently, I have also worked Part-time as an Adjunct Associate Faculty at Columbia University

Experience

Google

https://www.google.com

Software Engineer

May 2022 - Present

  • Working for the Google Ads product in the Audience Management team
  • Building a platform to help advertisers reach the right users across various advertising products
  • Developing a robust framework for using first-party data as the advertising industry is moving towards a cookieless world.

Oracle

https://www.oracle.com

Senior Software Engineer

March 2020 - May 2022

  • Working for the Oracle Transportation Management (OTM) Team at Oracle.
  • Worked on the cross-functional teams across the product
  • Currently Building Dashboards and Visualization components, as well as developing machine learning algorithms for the product

Columbia University

https://www.columbia.edu

Adjunct Associate Faculty

May 2020 - Jan 2021

  • Hired for the Course “Data Science Consulting” for two semesters.
  • Directed and helped Students develop Data Science capstone Projects in a Consulting-based environment.

Edlab, Teachers College, Columbia University

https://edlab.tc.columbia.edu

Data Science Researcher

June 2019 - Dec 2019

  • Worked on a Multi-platform recommendation system for various Teacher’s college library resources for enhancing Self Directed Learning (SDL)
  • Own entire data pipeline and perform end to end data engineering
  • Involves Development and research on Search and Recommendation systems, Social network analysis and Topic modelling
  • Also worked on auto-tagging of documents and designed a python package for research metadata retrieval
  • Technologies: Apache PredictionIO, MongoDB, EZProxy, Docker, Python, SQL

Columbia University

https://datascience.columbia.edu/students-recognized-excellence-during-reception-dsi-graduates

Graduate Teaching Assistant

Jan 2019 - May 2019

  • Recieved Award for Excellence in Course Assistantship
  • Teaching Assistant for the course “Applied Data Science”
  • Course focuses on the practice of Data Science in the field of Educational Assessment, Health Care, Marketing, Advertising and Social Media
  • Course deals with skills like Data analysis, Machine learning, Featuring Engineering and Interpreting & Communicating results.
  • Responsibilities include evaluating assignments & projects and providing guidance to graduate students.

Micro Focus

https://www.microfocus.com

Software Engineer

Sept 2017 - June 2018

  • Designed, tested and automated APIs for providing data analytics insights into software security compliance and risk data for a datacenter automation Software
  • Implemented the software using python programing on kubernetes technology
  • Won 2nd place in a site wide hackathon conducted by Hewlett Packard Enterprise by designing a ‘Docker and Kubernetes log monitoring Solution’
  • Technologies: Kubernetes, Docker, Python

Hewlett Packard Enterprise

https://www.hpe.com

Software Engineer

Sept 2016 - Sept 2017

  • Spin merged into Micro Focus in September 2017

Education

Columbia University, Data Science Institute

Master of Science - Data Science

2018 - 2019

B. V. B. College of Engineering and Technology (Visvesvaraya Technological University)

Bachelor of Engineering - Information Science and Engineering

2012 - 2016

Projects

Editorial Classifier

A Bloomberg Sponsored Capstone Project to classify news articles

  • Worked with unstructured news sources to build classifiers to identify editorial content for Bloomberg Terminal.
  • Developed traditional classification algorithms with text based features like n-grams, Parts of speech, Named entities etc.
  • Trained Natural Language Processing based Bidirectional LSTM, BERT, and XLNET Models for classification using Python.
  • Languages: Python(Pytorch, TensorFlow, spacy), R(ggplot)

Educational Recommendation System

Recommendation system to recommend digital resources to students

  • Developed a multi-platform database pipeline for physical and digital user behavior, and research content using AWS.
  • Built and evaluated a hybrid recommendation system on multiple digital learning platforms
  • Technologies: Python (nltk, predictionIO), SQL, MongoDB

Does Climate Change and Natural Resource Storage Cause Conflict?

A "Data Science and Public Policy" Project to Analyze Conflicts in Senegal

  • Find correlation between factors such as Rainfall, Temperature, food prices and Bio-mass and Conflicts in Senegal
  • Used Regression, SVM, Decision Trees and Time Series based models to predict Conflict. Link
  • Built a reporting Dashboard that can be used by policymakers to take effective decisions to avoid natural resources shortages and conflicts.

How America Flies

Creating Insightful Visualizations on Airline Performance Data

  • Performed Exploratory data analysis (EDA) and answered questions of interest on Airline On-Time Performance data for US
  • Conducted the EDA using R’s tidyverse packages.
  • Designed a Visualization tool in JavaScript and D3 to find flight delays between top airports.
  • Tool Link

Beer Recommendation System

A Recommendation system for recommendation of Beer

  • Built neighborhood based and model based collaborative filtering algorithms for beer recommendation
  • Developed a Content based recommendation system using NLP on beer review-text
  • Tackled recommendation concerns such as cold- start, variety, serendipity and accuracy
  • Developed the project in python using surprise, NL toolkit and scikit-learn packages

Projects in Data Science and Public Policy

Mini Data Science Projects in Public Policy Context

  • Russian Twitter trolls and 2016 U. S. Election – Analyzing FiveThirtyEight’s 3 million tweets dataset Link
  • Evaluating Northpointe Inc’s Compass system recidivism predictor Link
  • Predicting Economic factors related to Poverty using Night Lights dataset Link

Research Project in Educational Analytics

Evaluation and Validation of Problem Solving and Thinking Skills based on Student Academic Performance

  • Implemented a modified K-means clustering algorithm on a Weighted Skill Score (WSS) to evaluate Problem solving and thinking skills of a student based on academic performance
  • Analysis was done using R Programing
  • Presented and published at IEEE International Conference on Recent Trends in Electronics, Information & Communication Technology (RTEICT) 2017, Link