---
title: "Student Assistant at Escuela de Ingeniería Informática UCAB"
company: "Escuela de Ingeniería Informática UCAB"
position: "Student Assistant"
period: "February 2020"
contract: "Seasonal"
description: "I was responsible for making demand projections for the School of Computer Engineering and the School of Civil Engineering using Microsoft Access."
website: "https://ingenieria.ucab.edu.ve/informatica/"
url: "https://rolandoandrade.me/career/student-assistant-ucab/"
author: "Rolando Andrade"
---

# Student Assistant at Escuela de Ingeniería Informática UCAB

> I was responsible for making demand projections for the School of Computer Engineering and the School of Civil Engineering using Microsoft Access.

Period: February 2020. Contract: Seasonal. Company website: https://ingenieria.ucab.edu.ve/informatica/

# Predicting how many students come next

In February 2020, I worked as a student assistant and data scientist at my university, Andrés Bello Catholic University in Caracas. I worked for its engineering schools. The department needed help with decisions about hiring professors and planning classes.

They had thousands of students, but they made those decisions based on guesses instead of data. They needed to know how many students would take each class next semester, so they could plan better.

My job was to analyze the data and help them predict that number. These predictions are called demand projections. I made them for two schools: the School of Computer Engineering and the School of Civil Engineering.

## Cleaning messy data

The data was not clean. It lived in an old Microsoft Access database that had been in use for years. Access is a desktop database program from Microsoft.

The data had a lot of problems. Names were spelled wrong. Dates came in different formats. Some records were duplicated.

I spent weeks cleaning it. I wrote database queries to fix errors and organize the information. A query is a question you ask a database, like "show me every student who took this class last year."

## A model from past trends

Then I built a model that looked at past trends. It predicted how many students would take each class.

## Better plans for each semester

My reports helped the department make better decisions.

They could open new sections of popular classes before students registered. They knew where they needed to hire more professors. They avoided classrooms that were too full or too empty.

This job taught me that working with data is mostly about cleaning and organizing it, until you can see the patterns.
