Skip to content

About us

We are an AI technology and data intelligence startup company focused on the job and education market.

Our purpose is to democratize and equalize the job and education supply and demand in all social levels.

Our Project

Jobzi has an unique AI technology to collect, deduplicate and normalize job & education data to build predictive models

We also use Machine Learning techniques to build explainable models to predict connectivity at schools. To find the model(s) that will be implemented in our service we will invest in feature selection/engineering and we will also carry-on experiments with a number of possibilities ranging from classical models (Learners) as Decision Trees and SVM's to more recent ones, including Gradient Boosting Methods and Deep Learning.

Input data

  • Job Openings
  • Economic Index
  • Telecommunications Data
  • Education & School Data
  • Development Index
  • Research Institutes
  • Government Reports

How we are helping

Feature Through Interactive Map Rxplorer

Explore schools connectivity data, schools connectivity prediction and educational performance given to connectivity generated by our models on an interactive map.

Feature Through Analysis Tool For Prediction Models

A web tool that consists of two services to address the crucial role of Internet Connectivity in schools. This application results from our partnership with UNICEF, and we are proud of the outcome.

To ensure the quality of our services, we collected extensive data from various Brazilian government sources and studied their impact on the quality of the services developed. We are excited about the fruits of this challenging endeavor, which we believe will make a difference.

However, as this is the first version of our tool, we welcome everyone interested in contributing to improving this project.

Schools Connectivity Classifier

The first service is an Internet Availability Predictor that aims to provide a solution for schools lacking internet availability information. It helps policymakers plan and allocate resources more effectively by estimating the likelihood of schools having reliable internet access using demographic and geographic data provided by the user.

The statics models implemented for the study are mainly: Decision Tree, Random Forest and XGboost.

For more detail about the implementation, training and optimization: https://github.com/Jobzi-Artificial-Intelligence/ziconnect-backend-fastapi/blob/master/fastapi/services/internetConnectivityService.py

To run the model we have defined two templates to be loaded:

Employability Impact Classifier

The second service measures the employability impact of providing internet access by comparing data on employment rates and job opportunities between municipalities where school internet access has been improved or stagnated over the years. This information helps policymakers and educators understand the importance of reliable internet access for students' future success. Our study and results treated through the project is related more to the exploratory analysis and the correlation between connectivity and employability.

Allows users to download our data, execute our models, visualize the results/analysis and also download them.

For more detail about the implementation, training and optimization: https://github.com/Jobzi-Artificial-Intelligence/ziconnect-backend-fastapi/blob/master/fastapi/services/employabilityImpactService.py

To run the model we have defined two templates to be loaded:

Open Data License & Public DataSets Integration

Following the availability and distribution Open Data license from Brazilian government, the match for the Open Data license is: PDDL v1.0

GPDR Compliance

In agreement and evidence with this topic, the datasets applicable in this project doesn't apply to personally idenifying data. Additionally, the Data Providers follow de Public Open Data Source License: https://www.gov.br/governodigital/pt-br/dados-abertos.

The offical website https://unicef.jobzi.com/data-source-reference we have the GPDR Compliance available for our users and development and data science communities.

Public Databases

The project is collecting and using public data for the map explorer and precdition models, considering the main private and public institutions:

- Brazilian Government Instituions
    - INEP (Schools dataset): https://www.gov.br/inep/pt-br/acesso-a-informacao/dados-abertos/microdados/censo-escolar
    - IBGE (Brazil demographic dataset): https://www.ibge.gov.br/cidades-e-estados 
    - ENEM (Schools and students micro dataset): https://www.gov.br/inep/pt-br/acesso-a-informacao/dados-abertos/microdados/enem-por-escola 
    - IDEB (Schools dataset): https://basedosdados.org/dataset/br-inep-ideb?bdm_table=escola

- Private Institution
    - GIGA: (Brazil schools and demographic dataset): https://giga.global/

Open Data Formats

Considering the initial raw data structure performed, the platform has available the following files:

- https://www.ibge.gov.br/
- https://www.ibge.gov.br/estatisticas/economicas/comercio/9016-estatisticas-do-cadastro-central-de-empresas.html?=&t=resultados
- https://www.gov.br/inep/pt-br/areas-de-atuacao/avaliacao-e-exames-educacionais/saeb
- https://basedosdados.org/dataset/br-inep-ideb?bdm_table=escola
- https://www.gov.br/inep/pt-br/acesso-a-informacao/dados-abertos/microdados/enem-por-escola
- https://www.gov.br/inep/pt-br/acesso-a-informacao/dados-abertos/microdados/censo-escolar
- https://www.ibge.gov.br/cidades-e-estados