Avatar

Omar Pardo Gomez

Data Scientist

Who am I?

Having a background in Applied Mathematics, I have spent the last seven years understanding human behavior through data. Currently, I’m a Data Scientist working for Facebook, and previously interned at Airbnb. I got my MS in Data Science from Columbia University, and I also worked in the marketing and banking industries as a Statistician.

I’m passionate about democratizing knowledge in a way that is friendly for everyone, no matter their background. Consequently, I’m the host of datos netos, a podcast in Spanish in which I talk with experts about social phenomena from a quantitative perspective.

Interests

  • Human behavior
  • Bayesian Statistics
  • Machine Learning
  • Improving decision-making
  • Democratization of knowledge

Education

  • MS in Data Science, 2019

    Columbia University

  • BS in Applied Mathematics, 2016

    Insituto Tecnologico Autonomo de Mexico (ITAM)

Projects

*

Atleta

OOP Python project which simulates a hypothetical retirement plan for NFL players

Black Lives Matter

NLP analysis that breaks down the topics and emotions contained in tweets about the BLM movement

dlmRStan

R Package that fits a Bayesian linear model in a time series context

GPDPQuantReg

R Package which fits a Bayesian and nonparametric quantile regression model

Inside The Office

Exploratory data analysis and visualizations about The Office (NBC), using tidyverse

KPIsHealth

Shiny app useful to explore trends in time series, designed for KPIs estimated from survey data

Soccer Learning

Predicted which team has the ball possession in a soccer game by utilizing CNNs and RNNs.

Wes Anderson Recommendation System

Recommendation system to find users likely to watch each of Wes Anderson’s films, using Factorization Machines

Experience

 
 
 
 
 

Data Scientist

Facebook

Feb 2020 – Present Menlo Park, CA
  • Joined the Data Science team working on Video Ranking.
 
 
 
 
 

Data Science Intern

Airbnb

Jun 2019 – Aug 2019 San Francisco, CA
  • Personalized our marketing strategy by training machine learning models that characterize the users landing via paid and organic channels.
  • Built time-series dashboards to verify if such characteristics remain throughout time.
  • Boosted the probability of booking by finding the sequences of channels with the best results, influencing the creation of holistic strategies.
  • Designed a novel customer-centric metric framework that is the first-ever to account for the possibility of users liking multiple channels.
 
 
 
 
 

Statistical Modeler

BBVA MX (Mexico’s largest bank)

Jul 2017 – Jul 2018 Mexico City, MX
  • Increased the number of investments after automating targeted offers, fitting PySpark models that estimate monthly expenses for each of our users.
  • Improved our clients’ finances by suggesting personalized actions, given their current balance and trends.
  • Lifted the bank’s profit by estimating the interest rate a given customer is willing to pay for a personal loan, using random forests.
 
 
 
 
 

Junior Data Scientist

Kantar (Market research firm)

Jun 2016 – Jun 2017 Mexico City, MX
  • Empowered our clients to figure out the current state and momentum of their KPIs, by building dynamic Shiny apps.
  • Optimized our clients’ media strategy by estimating the effect of different advertising channels on their metrics, utilizing self-developed R packages.
  • Stopped our clients from advertising on TV shows that their current (and future) best customers don’t watch, by uncovering age-driven consumption patterns.

Relevant Courses

Applied Deep Learning (Prof. Josh Gordon)

Applied Machine Learning (Prof. Andreas C. Müller)

Machine Learning for Data Science (Prof. John Paisley)

Personalization: Theory & Applications (Prof. Brett Vintch)

Algorithms for Data Science (Prof. Eleni Drinea)

Bayesian Statistics (Prof. Manuel Mendoza)

Skills

Python

Excellent

R

Excellent

SQL

Excellent

Git

Proficient

LaTeX

Proficient

PySpark

Familiar

Contact