Brian – React, Python, AWS

This client was a company dedicated to providing gender-based analysis to institutions and organizations, fostering equity and informed decision-making. It offered tools like a Retrieval-Augmented Generation (RAG) chatbot, combining data retrieval with AI to deliver precise and context-aware insights. These solutions enabled organizations to access actionable information efficiently.

• designed and implemented a Retrieval-Augmented Generation (RAG) chatbot, using Claude LLM, MongoDB vector search engine, Python and Node js;
• conducted maintenance of existing codebases and resolved bugs within established timelines.

A company focused on incentivizing farmers to adopt environmentally friendly practices that reduce CO2 emissions and minimize soil damage. Farmers earn credits for sustainable methods, which can be exchanged for money at the end of each campaign, while their certified products carry a seal signaling eco-conscious practices to consumers. By collecting detailed data on crops, planting methods, chemicals, and more, the company calculated CO2 emissions and evaluates the impact on soil health.

• implemented a Python-based model translator software, delivering increased flexibility and speed to the project that stores the data models, resulting in a 30% improvement in efficiency;
• optimized 2 processes within an AWS environment, using Amazon S3 and EC2 instances for efficient infrastructure management;
• created a new feature that reduced data processing time from two weeks to three days, using Python for the backend and JavaScript.
• fixed critical bugs in production in less than 6 hours;
• developed the automation of repetitive tasks with Bash scripts, minimizing non-development-related task downtime by at least 50%.

A company specialized in aggregating health insurance plans from carriers' websites and endpoints, structuring the data into an organized database. This information was presented through a user-friendly web interface, enabling brokers to search plans based on criteria like client numbers, age, location, and medical specialty. The platform provided brokers with streamlined access to insurance options, available via a subscription-based model.

• automated several processes, reducing a ten-minute file upload task to one minute and enabling the upload of 100 plans per minute using JavaScript and React;
• enhanced data automation solutions for multiple insurance brokers and their database management;
• increased weekly insurance plan uploads from 500 to 1000+ through automation, utilizing Python and React for development;
• implemented significant UI improvements, enhancing user experience and optimizing workflows.

This project was a software development and consulting company, offering tailored solutions across various industries. Its product portfolio included a cable car ticketing system, custom home banking systems for cooperatives, and payment networks for financial institutions. The company delivered innovative and reliable technology to meet diverse client needs.

• integrated payment service into websites using Java, facilitating online payments by embedding a payment gateway solution;
• solved architecture issues related to route calculation processes in a Java-based MVC architecture, reducing processing time from over an hour to 10 minutes and eliminating system freeze;
• implemented UI components using Java frameworks, enhancing the user interface of the application with rich and interactive elements;
• created a Home banking with basic operations that included authentication, security, account creation, money transfers, and reporting, using Node.js.

Galindo consulting Group is an organization in charge of a portfolio of different clients and industries, that includes: banking, insurance, fintech, personal websites, start-up prototypes and more.

• created a trading bot, using React for creating the UI that let the end user configure the trading strategies and analiticals indicators.
• used Python to run the logic for the final signals, sell, buy of no-action.
• a fully automated process was created that reduced the time to upload information from multiple rweb sources to the client's server from 1 week to 5 minutes.