Most companies have laborious tasks that demand a long time. One of the purposes of data science or analysis is to automate those kinds of processes, reducing expenses and allowing focus on other important aspects for companies. In this project, we drastically reduced the time spent on the creation of reports for a company in charge of occupational health and safety at work.
TARGETS
PROJECT DURATION
Approach
The first step was to collect all the reports generated between 2022 and 2023 for all the construction companies, using APIs to extract all the information saved, cleaning the data, and removing empty entries. Failures in the database were also verified, taking into account that there were over 3600 reports, each with around 20 stratified questions and 5 possible answers.
Once the previous task was completed, the next step was to work on and manipulate the correct columns of information, as we had a vast database with redundant or unnecessary columns that were not useful for our target.
Continuing with the process, it was required to group the information by customer code (the reference number for every construction company) and the associated site for every construction company. The purpose of this was to show and track the progress of each construction work.
Later counting with many levels of risk (Low, high, Medium, Stop work) there are by work construction and by category, all of these along of the time
Another really useful tool is the availability of a display by category, focusing on each construction company and the number of levels of risk at any given time