Data Lab

The Datalab team improves decision making and operational efficiency through data analysis and predictive modeling, developing advanced technology solutions to optimize operations and ensure information quality.

Mission

Support the decision making process to ALTO’s internal and external clients, through the analysis of information, design and application of predictive and prescriptive models, reaching measurable results for our business and thus support intelligent business decision making. The purpose is to contribute to ALTO’s scientific and technological development, leading research and technological development based on information that generates operational efficiency.

Formation of the team

Interdisciplinary group with extensive knowledge in information analysis methodologies and development of mathematical models, both predictive and prescriptive.

Datalab – project summary This information is in Spanish, if you require translation you must submit the request at the following link: https://soporte-ti.atlassian.net/servicedesk/customer/portal/14/group/83/create/868

Relevant experience and developments of the DataLab team

The team analyzes information to generate special projects and raise opportunities to improve operations and generate efficiency.

Information Analysis

  • Perform descriptive analysis from large volumes of data.
  • Design indicators to measure the quality of information.
  • Find insights from data analysis.
  • Analyze information from non-traditional sources, such as video, audio and text.

Modeling

  • Deep Learning models are developed for image and text processing and analysis.
  • Develop machine learning models for the prevention of asset losses due to shrinkage.
  • Develop optimization models for scheduling schedules and personnel tasks (lawyers, call center personnel).
  • Development of Machine Learning models for pattern identification.

Quality of information

  • Implementation of imputation methodologies and information quality.
  • Automatic collection of structured and unstructured data from external sources.

Awards and track record

2019 – Challenge Viva Colombia – University of Los Andes, Bogotá – Colombia

The challenge consisted of forecasting the horizontal and vertical visibility of the airports associated with Viva Colombia. Visibility is a critical aspect, since, if this variable is not known, the result is cancelled flights or delayed flights. To solve this problem we developed a solution that involves direct connection to web sites that provide information about key variables that allow us to make such predictions, for example pressure, temperature, etc. Then this information went through a quality process, where it is debugged, cleaned, and the information is narrowed down. Subsequently, it is used to feed deep learning models. These models not only make predictions for different times, but also activate alerts that inform the VIVA COLOMBIA network. This process is coupled in a single web solution, which in real time is capable of describing, forecasting and alerting, thus fulfilling the analytical integration for decision making.

Trayectoria

Research Agreements – University of Los Andes

  • Analytical model for the identification of anomalies in the behavior of electric energy service users, Master’s Degree in Analytical Intelligence for Decision Making. Bogota, Currently 2020
  • Significant geo-referenced factors in the explanation of monetary losses due to shoplifting in Mexican stores. Master in Analytical Intelligence for Decision Making. Bogotá, December 2019.
  • Prediction and Characterization of Property Crime Losses in Stores in the United States. Master in Industrial Engineering. Bogota, June 2019

Publications

Publications High

Research projects

Image processing

Information quality management

Highlights

Image processing applications

Curriculum Datalab

This information is in Spanish, if you require translation you must submit the request at the following link: https://soporte-ti.atlassian.net/servicedesk/customer/portal/14/group/83/create/868

How much did you learn?

In the following link you will find questions related to the above explained.

Data Lab Questionnaire This form is in Spanish, if you require translation you must submit the request at the following link: https://soporte-ti.atlassian.net/servicedesk/customer/portal/14/group/83/create/868

CLASSIFICATION: INTERNAL USE
Any printing or digital copy of the document is invalid. The user of the document must ensure that it is current before using it. This can be checked with the person responsible for the ISMS.