Sebastian Echeverri

Data Analyst

My Areas of Expertise

Data Analyst with over 2 years of experience of extracting data and translating it to actionable insights for multiple companies.

Technical Skills
  • Python
  • Data Cleaning, Managment
  • SQL, Mongo DB
  • Javascript
  • HTML
  • CSS
  • VBA Macros(Excel)
  • R Programming Language
Downlaod CV

Machine Learning

Creating predicitve models by using both python and R. Learned different modules in both languages to create the most accurate models possible. Proficient in using random forest models, cross validation techniques, logistical regressino models and XGB regression models mostly in python.

Webscraping

Experienced in webscraping by using python modules such as beautiful soup, splinter and selenium. Used these skills to extract, and organize competitors techniques.

Extract, Transform and Load

Proficient using SQL and MONGO DB in the ETL process and analyzing and cleaning the data with python.

Education

Data Analytics Bootcamp

University of Miami
2019-2020

B.A. in Industrial Engineering From

University of Miami
2015-2019

Work Experience

Excell LTDA

Supply Chain and Price Analyst
May 2019 – July 2019
Responsibility :
  • Updated the KPI’s for the warehouse operations to manage and capture. Updated KPI’s such as inventory turnover, cost of carrying inventory, receiving efficiency, number of write downs, inventory accuracy, and workforce utilization.
  • Created a predictive model using R to forecast demand with an 85% accuracy rate. Previously the company had no such model in use. It helped them lower their inventory cost by 15% and allowed them to fufill 10% more of their orders.

Ryder(Transport Company)

Capstone Project with University of Miami

Jan 2019 – May 2019
Responsibility :
  • Created a model which can determine the optimal holding period for Ryder's rental vehicles.
  • Analyzed and cleaned the data by using a combination of Excel and R programming language, which was necessary to create a linear predicitve model of the cumalitive margin at any month of each vehicle

Terra Group (Real Estate Development)

Financial Modeling

August 2018– May 2019
Responsibility :
  • Built financial models for potential projects, which helped filter which projects to consider.
  • Built predictive models for future cashflows based on population growth, income per capita in the area, and current average cap rate. This model was 10% more accurate than the previous model which they were utlizing.

Recent Portfolio

  • All Categories
  • Webscraping
  • ETL
  • Data Analytics
img04
Weather Analytics
Data Analytics
img04
Mars Webscraping
Webscraping
img04
Titanic: Machine Learning from Disaster
Analytics/Machine Learning
img04
Matplotlib Challenge
Data Analysis
img04
Belly Button Biodiversity Project
ETL

Matplotlib Challenge

The purpose of this project was to analyze how the weather changes as you get closer to the equator. To accomplish this analysis data was pulled from the OpenWeatherMap API to assemble a dataset on over 500 cities.

After assembling the dataset, we used Matplotlib to plot various aspects of weather vs. latitude. Factors we looked at included: temperatue, cloudiness, wind speed and humidity. This site provides the source data and visualizations created as part of the analysis as well as explanations and descriptions of any trends and correlations witnessed.

  • python
  • html
  • css
  • Javascript
  • Matplotlib
  • Bootstrap
Link of the website

Mars Webscraping

The purpose of this project was to analyze how the weather changes as you get closer to the equator. To accomplish this analysis data was pulled from the OpenWeatherMap API to assemble a dataset on over 500 cities.

After assembling the dataset, we used Matplotlib to plot various aspects of weather vs. latitude. Factors we looked at included: temperatue, cloudiness, wind speed and humidity. This site provides the source data and visualizations created as part of the analysis as well as explanations and descriptions of any trends and correlations witnessed.

  • python
  • html
  • css
  • Javascript
  • Matplotlib
  • Bootstrap
Link of the website

Titanic: Machine Learning from Disaster

The purpose of this project was to create a predictive model by using machine learning that answers the question of what sorts of people were more likely to survive. This was done by using the passanger data that was given in the competition such as age, gender,socio-economic class, and number of family members on board. The data to train the model and to test the model was given in a csv folder. By using data cleaning all of the nan values were replaced and all of the columns could be used to imrpove the model.

A random forest regressor model was chosen and then was tested using the Test data and was given %78 accuaracy. Below the link to the github repository will be posted

  • Python
  • Sklearn
  • Pandas
  • Data Cleaning
Link of the github repository

Mars Webscraping

The purpose fo this project was to build a web appliation which scrapes multiple websites for data related to Mars. Some examples of the data are the current weather one mars, mars weather twitter acount's last tweet, and pictures of the four hemispheres of Mars

The web application runs using html, javascript, css, and python. Therefore the website which I deployed on Github's services cannot run but below I will attach a link to my source code.

  • python
  • html
  • css
  • Javascript
  • splinter
  • Twitter API
  • Bootstrap
Link to the source code

Belly Button Biodiversity Project

The purpose of this project was to build an interactive dashboarrd board to explore the belly button Biodiversity dataset. This dataset catalogs the microbes that colonize human belly buttons.

The datset reveals that small handful of microbial specias(called taxonomic units,OTUs in the study) were present in more than 70% of people in the study while the rest were relatively rare.

  • python
  • D3
  • json
  • Plotly
  • html
  • css
  • Javascript
  • Bootstrap
Link of the dashboard

Weather Data Visualization

The purpose of this project was to analyze how the weather changes as you get closer to the equator. To accomplish this analysis data was pulled from the OpenWeatherMap API to assemble a dataset on over 500 cities.

After assembling the dataset, we used Matplotlib to plot various aspects of weather vs. latitude. Factors we looked at included: temperatue, cloudiness, wind speed and humidity. This site provides the source data and visualizations created as part of the analysis as well as explanations and descriptions of any trends and correlations witnessed.

  • python
  • html
  • css
  • Javascript
  • Matplotlib
  • Bootstrap
Link of the website