🌎 Let's change the world 🌏
Software Engineer, Freelance Consultant, Indie Hacker, Professor, and more...

github

projects

productHunt

dribbble

leetcode

youtube

Rolando Andrade  • Peerlist

José Gregorio is an AI-powered medical assistant designed to help Venezuelans access medical assistance promptly and efficiently. Named in honor of the beloved Venezuelan doctor José Gregorio Hernández, who is revered for his dedication to the sick and needy, this AI embodies the same spirit of service. AI José Gregorio uses LLAMA2, an semi-open source LLM released by Meta, and was trained on medalpaca/medical_meadow_medqa dataset, making it capable of understanding a wide range of medical inquiries and providing useful, potentially life-saving, information. The original model (medllama2) was able to pass the US Medical Licensing Exam. It was implemented using React-Native (Expo), Express.js, and Python

AI José Gregorio

As the developer responsible for the customer portal at Daycohost, one of the largest data centers in Venezuela, I am proud to have developed a system that has been widely adopted by a growing number of companies and users. By 2022, the customer portal I created was being used by over 100 companies and 500 users, which is a testament to its success and effectiveness in meeting the needs of customers in the data center industry. The customer portal I developed for Daycohost was designed to provide users with a convenient and user-friendly platform for managing their accounts, accessing their data, and interacting with the company's support team. This included features such as the ability to view and manage billing information, request technical support, and access real-time monitoring data for their servers. Through my work on the customer portal, I was able to help Daycohost improve its customer experience and streamline its operations. By providing customers with a reliable and efficient platform for ticket management and other business processes, we were able to improve customer satisfaction and reduce support requests. I believe that it has made a significant impact on the company's success and reputation in the data center industry.

Daycohost Customer Portal

Filecoin-js-signer is a pure Javascript/Typescript library that provides a complete implementation of the Payment Channel actor's methods, enabling the creation of atomic swaps and atomic loans in the Filecoin network. The library includes necessary methods to create, sign, and broadcast messages to send FIL, and interact with Filecoin's built-in actors such as Payment Channel and Multisig. The library was initially created as part of the Filecoin Loans Grant awarded by the Filecoin Foundation in April 2021. At the time of its creation, there were no fully functional signing libraries written in pure Javascript or Typescript, and this library aims to fill that gap. Additionally, a pure Javascript implementation of these methods can simplify the usage from web and mobile applications that interact with the Filecoin network. By providing a comprehensive and functional signing library, Filecoin-js-signer aims to promote broader adoption of the Filecoin network and enable developers to build more complex and secure applications on top of it.

filecoin-js-signer

General Simulation Framework (GSF) is a powerful tool for implementing simulation projects in a wide range of fields in Python. This framework is designed to offer a standard structure for simulation tools, allowing developers to reuse and extend components to generate a variety of simulation projects. By using GSF, developers can save time and effort by avoiding the need to learn multiple simulation processes and programming from scratch. When I was implementing some simulation tools I discovered a pattern among them, so I thought I could abstract the core features and implement a framework for it. One of the key benefits of GSF is its ability to streamline the simulation development process, allowing developers to create solutions more efficiently while maintaining a high level of quality and consistency. Contributions from the open source community are also welcomed, as they help to improve the framework and make it more accessible to a wider range of users.

General Simulation Framework

The Genetic Algorithm Framework is a small, lightweight framework designed to facilitate the handling of genetic algorithms. It provides a set of tools to define agents and their genetic functions, as well as the simulation functions necessary for the genetic algorithm to work. The framework is written in TypeScript and can be used with any programming language that can use JavaScript libraries. It was implemented after seeing that most of my Genetic Algorithm projects had the same scheme, so I decided to abstract that and implement a generic library for it. The components allow defining the genetic functions and the simulation runs everything on background so users do not have to care about implementing the genetic algorithm steps and connections

Genetic Algorithm Framework

IMG Processing is a powerful and flexible API that provides a wide range of image manipulation and analysis capabilities, allowing developers to integrate advanced image processing features into their applications with ease. Some of the features are: Image storage, image transformation, image adjustments, background removal, format conversion, watermarking, image generation, image classification.

IMG Processing

Me is a mobile application designed to help you set and achieve your goals while discovering new things you enjoy. The application allows you to set goals and track your progress towards achieving them, providing reminders and push notifications to keep you on track. The app was made with React-Native (Expo) and helps individuals who are looking to set and achieve personal goals while discovering new activities they enjoy.

In the digital age, we need digital documents. Scan Documents is created to digitize your world. Just take a picture of a page, book, or ID and the tool will attempt to restore the image by dewarwing, enhancing, and applying effects to the image, even offline!There are two ways to use Scan Documents.Normal users can use the free WebApplication to digitize their documents in a easy and fast way, even offline since no documents are sent to any server. On the other hand, developers have available a powerful subscription-based API that can use to automatize document digitization workflows like document scanning, color and perspective correction, format conversion, and more.I really enjoyed building this and I hope you can enjoy using as I do!

Scan Documents

Simulab is a web-based application that allows users to simulate discrete-event queue networks using Vue.js, Flask, Socket.IO, and the General Simulation Framework (GSF). The application provides a user-friendly interface for creating and running simulations. The General Simulation Framework is a Python library that provides a flexible and extensible platform for building simulation models. Simulab uses GSF to model the behavior of queue networks, including the arrival and departure of customers, queueing delays, and service times. The web interface of Simulab is built using Vue.js, a progressive JavaScript framework that makes it easy to develop dynamic user interfaces. The interface allows users to create simulation models by specifying the network topology, arrival and service distributions, and other simulation parameters. Flask is used as the web server framework for Simulab. Flask provides a simple and flexible architecture for handling requests and responses for the app. Socket.IO is used to enable real-time communication between the client-side web application and the server-side simulation engine. This allows the simulation results to be displayed in real-time as the simulation runs

Simulab

This is a module that provides an easy way to map objects from a template skeleton. It takes in a source object or an array of objects and a skeleton template, and maps the source according to the skeleton. The skeleton is a template that defines the structure of the resulting object. The motivation behind this project is to simplify the process of mapping objects from a source to a target structure. This is particularly useful when dealing with data from different sources that have different structures, or when transforming data for use in different parts of an application. By simplifying the process of mapping objects, the module can save developers time and reduce the likelihood of errors in their code.

skeleton-mapper

StepCode is a pseudo-code language that offers a user-friendly and intuitive approach to writing step-by-step instructions. Designed with simplicity and ease-of-use in mind, StepCode empowers both novice and experienced programmers to express complex algorithms and logic with remarkable clarity. Drawing inspiration from the syntax of PSeInt, a renowned pseudo-code language used in programming education, StepCode takes the concept further by introducing advanced features and a modernized framework. At the heart of this project lies the implementation of the StepCode programming language itself. Meticulously crafted using my expertise in compilers and parsers, the language boasts an elegant and concise syntax that enables programmers to express their ideas in a highly readable and comprehensible manner. By utilizing familiar keywords and constructs, StepCode ensures a smooth transition for individuals already accustomed to PSeInt or other similar pseudo-code languages. However, StepCode goes beyond its predecessors by incorporating additional functionality and flexibility, allowing for the seamless conversion of pseudo-code programs to various programming languages. To complement the StepCode language, I implemented a web-based editor named StepCode Editor. This powerful editor based on Codemirror (Monaco Editor formerly) serves as a versatile platform for writing, running, and testing StepCode programs across all major operating systems and devices. Its intuitive user interface provides a seamless coding experience, empowering programmers to focus on algorithmic logic rather than intricate technicalities. The StepCode Editor's compatibility with multiple platforms ensures a consistent and accessible coding environment, enabling programmers to work efficiently without any constraints.

StepCode & StepCode Editor

At Unbanked, we frequently had partners who wanted to use their own crypto tokens as a valid asset for spending in the real world via the Unbanked crypto card. However, the implementation of a new crypto token was a complex process due to the tight integration of blockchain capabilities with the system core. The token needed to be deposited, withdrawn, exchanged, etc., and each of these flows required multiple systems and subsystems that were tied to the blockchain assets. Upon observing this, I realized that the process of inserting a new asset took too long and was prone to errors. I decided to split the large subsystems into smaller ones, which allowed me to decouple the asset from the processes. This required me to create multiple contracts and interfaces at the code level, and it was a challenging task to decouple the pollers that listened to transactions on blockchains. However, once the decoupling was done, I had modular building blocks that could be added or removed without affecting the system as a whole. To take things further, I designed a plugin system for assets and blockchain listeners, which turned out to be a game changer. Previously, implementing a new asset on a supported blockchain took two weeks, but with the plugin system, I was able to create a process that could do it in just two minutes. This allowed me to lead the integration of multiple partners simultaneously in less than two weeks, which directly increased the B2B revenue to over 1.2 million and introduced new tokens to the B2C solutions, which positively impacted the revenue.

Rolando Andrade

🌎 Let's change the world 🌏 Software Engineer, Freelance Consultant, Indie Hacker, Professor, and more...