DOMAS is a device that helps blind people to feel the shapes and the position of a certain object in front of them. This is an alternative to the walking stick.

DOMAS encodes shapes, figures, and positions, in real-time, of objects shown by means of a video camera and decodes them in the form of vibration, by means of an array of micromotors.

I was in charge of the programming and electronic design of the device using Python, C ++, and Proteus.

You can find more information about this project in the following PDF link on page 128, Chapter 9: "2D object shape and position identification wizard for visually impaired people".

ISBN: 978-958-52097-0-1

PDF

CLI+ is a meteorology station that monitors the weather in a given place. This gadget is auspicious to measure the weather variables like Temperature, Pressure, Humidity, etc. in farms, paragliding ditches, crops, or whatever place where you need to be aware of the climate.

The device stores the last 24 hours of meteorological measurements and sends, every 2 minutes, the current average of each meteorological variable to an API using C ++ and GSM.

I was in charge of developing the service which stores and sends the data from CLI+ to an API.

In the next link, you will be able to find one of the potentials of CLI+: The downpour of May 26

I participated in the project "Study and application of the NLP2000 through the student's guide from the Nanoprofesor laboratory" whose mission was to look for techniques of nanomanufacturing for developing emerging technologies such as sensors, inks, solutions, inks, substrates, tips, and software management using the NLP2000 the area of Biotechnology and Nanotechnology at TecnoParque.

I was in charge of discribing the distintics features the NLP2000 equipment has when making nanomanufacturing at TecnoParque.

The main conclusion of this project was that is possible to make a microcircuit with the NLP2000 using the lithography scanning method.

I participated in the project "Scanning lithography as a technique for the creation of microsensors prototypes" whose mission was to research what are the best pattern designs and techniques of lithography for creating a basic microsensor prototype in the Biotechnology and Nanotechnology area at TecnoParque.

I was in charge of setting up the best lithography techniques for use in NLP 2000.

The main conclusion of this project was in order to create microsensors prototypes organic inks are required.

EmailScrapper is a service that helps Torre's recruiters to contact talents through a bot. Its main objective is to automate the process to instruct job seekers to create an account in Torre.co and so participate in the process of a given job offer.

I was in charge of the backend of the service. Here I had to program the Bot actions which were: get the talents' information from the given job board who apply to the Torre jobs, contact the talents via email, and show some stats of the talents to Torre's recruiters.

This service was developed using Nodejs, AWS, MySQL, Docker, Express, and React. Its architecture was based on Microservices.

GitHub

Airbnb Clone is a service, similar to Airbnb, where you can rent accommodations with the best amenities. Its main objective is to facilitate hire a room, or a house in a given place.

I participated in this project developing the user interface, creating the API, and setting the production deployment for this app.

This service was developed with Python, Flask, JavaScript, CSS, MySQL, HTML, Jinja2, Puppet and Nginx.

GitHub

Blur filter is an image filter that uses the convolution method for making an image loses sharpness. This filter uses a Gaussian window to blur the given image and uses multithreading to reduce computing time.

I implemented this filter using pthread library in C language.

GitHub

People identifier is a Deep Learning service that identifies people in images. This service makes a bounding box around the people found in the image and it shows the probability that there is a person in that box.

I developed this service using Keras and TensorFlow. The Deep Learning method for identifying the people was Convolutional Neural Networks using YOLO (You Only Look Once) algorithm.

GitHub

Art shell is a command-line interpreter that executes commands read from standard input in Linux. It handles multiple program execution, input/output redirections, and signals.

Art shell is based on "sh" command-line interpreter in Linux and was made 100% in C language.

I led this project and I was in charge of handling the programs execution and the input/output redirections.

GitHub

Numbers Recognition API is a Web service that integrates Machine Learning to digitize numbers printed in an image. For images that have a clear background and the printed number is dark, the accuracy of the service is greater than 97%.

I developed this service using Flask, Keras, and TensorFlow. The machine Learning method for recognizing the numbers was Convolutional Neural Networks.

GitHub

Music creator is a Deep Learning service which can create Jazz music. This service generates musical notes improvised, but taking into account previously created musical notes.

I developed this service using Keras and TensorFlow. The Deep Learning method for creating the Jazz music was Recurrent Neural Networks using LSTM (Long Short-Term Memory) algorithm.

GitHub