Matías Vásquez Orrego
Data Scientist at Almacén Gurú- Claim this Profile
Click to upgrade to our gold package
for the full feature experience.
Topline Score
Bio
Credentials
-
Getting Started with Power BI Desktop
CourseraJul, 2022- Nov, 2024 -
Google Data Studio Complete Beginners to Advanced Tutorial
UdemyJul, 2022- Nov, 2024 -
Neural Networks and Deep Learning
CourseraJul, 2022- Nov, 2024 -
Sequences, Time Series and Prediction
CourseraJul, 2022- Nov, 2024 -
Advanced Learning Algorithms
CourseraJun, 2022- Nov, 2024 -
Sequence Models
CourseraJun, 2022- Nov, 2024 -
Supervised Machine Learning: Regression and Classification
CourseraJun, 2022- Nov, 2024 -
Curso de Fundamentos de Bases de Datos
PlatziJan, 2021- Nov, 2024 -
Curso Profesional de Git y GitHub
PlatziDec, 2020- Nov, 2024 -
Curso de Web Scraping: Extracción de Datos en la Web
PlatziFeb, 2020- Nov, 2024
Experience
-
Almacén Gurú
-
Chile
-
IT Services and IT Consulting
-
1 - 100 Employee
-
Data Scientist
-
Aug 2022 - Present
The role involved working with labelers, data analysts and sales personnel to extract, process, and sold traditional markets transactional data per category to important supplier companies. ● Achived an 80% reduction in labeling total time from 2 years to 4.8 months of labeling by scaling up the labeling process ▹Developed a classification model with recurrent neural networks and deep learning to classify untagged products for each category using tensorflow ▹Reduced the number of products required to train the classification model using Pareto analysis and K-means algorithms ▹Implemented the algorithms and the model using AWS Sagemaker, ECS, Lambda, EventBridge, and python ● Improved the labeling tool based on Google Sheets with AppScript. Fixed the data losses that the tool had and optimized loading times from hours to minutes ▹Implemented periodic meetings with the labeling team in order to explain doubts about the tool and to raise errors ▹Created historical tables in the MySQL database to record each change made in the production tables through triggers ▹Optimized queries and created indexes in the MySQL database to reduce loading times ● Created on my own initiative a mail alert system for data extraction failures. This reduced delays in client deliveries by identifying issues in advance. ▹Developed configurable t-test automation using Scikit-learn and email alert service using the Google Gmail API in Python. ▹Deployed the system using Docker, AWS ECS and AWS Lambda. Show less
-
-
-
Yom Mercado
-
Chile
-
Software Development
-
1 - 100 Employee
-
Data Scientist
-
Jul 2021 - Jul 2022
The role involved collaborating with different areas to identify opportunities for using data-driven decision-making to improve team performance. ● Created business intelligence PostgreSQL database and dashboards that provided stakeholders with valuable insights and enabled informed decision-making across the organization ▹Conducted interviews with operations, sales, and finance teams to understand their data needs ▹Designed datamarts to address the identified questions ▹Developed ETL processes to load data into the database using python packages pandas, pymongo, sqlalchemy, and airflow ▹Deployed the ETL processes with Docker, CircleCI, AWS ECS, AWS Batch and AWS Eventbridge ▹Created dashboards using Data Studio to visualize the data and provide daily insights ● Achieved a 6% increase in sales margin by creating a simple pricing application. This application simplified the pricing process, resulting in a more efficient workflow and improved profitability ▹Developed a process for reading prices from purchase invoices, and utilized multiple price scrappers to compare our purchase prices with our selling prices in a structured manner, using Python packages pandas and requests ▹Deployed the system using Docker, CircleCI, AWS ECS, AWS Batch, and AWS Eventbridge. A prototype front-end was implemented in Google Sheets ● Achieved a reduction in losses due to expired products and fulfillment issues by forecasting demand for each product and maintaining a constant stock rotation ▹Developed a demand forecasting model using ARIMA, including a loop for reducing stock based on current inventory levels. Automated purchases from suppliers were also implemented using Python requests through suppliers' web API ● Conducted a thorough analysis of operational metrics and their correlation with customer churn rates to determine areas of focus for improvement. Based on the findings, we were able to implement targeted strategies to increase client retention Show less
-
-
-
Unitti
-
Chile
-
Information Technology & Services
-
1 - 100 Employee
-
Data Scientist Intern
-
Feb 2021 - Jun 2021
The role involved collaborating with different areas to identify opportunities to improve the operational process. ● Achieved the automation of multiple recurring tasks requested by customers via email ▹Developed an Extreme Gradient Boosting (XGBoost) model to classify the requested tasks ▹Automated tasks such as creating backup databases, generating new AWS EC2 instances, etc ● Assisted in identifying opportunities for optimizing server applications by implementing a low-cost server monitoring platform based on open-source tools ▹Implemented the platform using Prometheus, Grafana, and Loki in Docker containers within an AWS EC2 instance Show less
-
-
-
Komatsu
-
Japan
-
Machinery Manufacturing
-
700 & Above Employee
-
Project and Innovation Engineer
-
Apr 2019 - Mar 2020
The role involved collaborating with various teams to develop innovative products aimed at enhancing Komatsu's operations.● Developed algorithms that detected stationary conditions in the electrical signals of operating electrical motors. These algorithms were crucial in the detecting prominent faults system of the mining truck● led the purchase of a real-time emulator for the control system of the 930 mining truck series, and collaborated on the development of system simulations. This system was critical for improvement research and maintenance of control cards Show less
-
-
Engineer Intern
-
Apr 2018 - Mar 2019
The role focused on studying the inverter card for the electrical motors of the 930-E4 mining truck, which formed the basis of my university thesis.
-
-
Engineer Intern
-
Dec 2017 - Mar 2018
The role involved programming PLCs to replace the relay circuit on an electrical motor test bench developed by the innovation team.
-
-
Education
-
2012 - 2019
Universidad de Santiago de Chile
Electrical Engineering -
1997 - 2011
Saint Gabriel´s School
Basic and Secondary Education
Community
