Edgar – JavaScript, Node.js, React

A national school network that serves about 1000 students across their different campuses, they needed a new tech stack to handle their student's needs.

• Implemented google classroom for the entire organization (server configuration)
• Renewed their websites
• Created a custom CRM
• Created a custom LMS
• Implemented a fully managed network system

Edgar designed and developed an AI music curation platform for a company offering music-based therapeutic services for the elderly. Previously, curators had to manually browse and compile playlists, so Edgar built a chatgpt-like interface that automated and streamlined the process. The app featured a three-column layout: a conversation list, an interactive chat (powered by Gemini and AXL-LLM), and a Figma-style React Flow canvas for organizing playlists. Users could chat naturally with the AI to discover songs (e.g., “top 50 in the US” or “tracks similar to Ed Sheeran”), while the backend used interpreters to translate requests into structured API queries against the private music service. The results were displayed in an intuitive UI, allowing curators to drag and drop songs onto a visual canvas, create notes or themed playlists, and ingest them directly into the company’s system.

• Completed the MVP for the product;
• Successfully implemented a network of interpreters (AI decision makers) that correctly retrieved data as expected;
• Integrated several providers into a cohesive platform (Gemini, Open AI, Billboard API, private music service API, Spotify API, clients API).

A platform that leverages AI voice and chatbot services to automate wellness checks and medication reminders for elderly care organizations in the U.S. and U.K. Instead of manually calling each client, agencies can deploy these fully automated voice agents, built using APIs such as VAPI Elevenlabs Twilio, and AWS, to interact naturally with users, record responses, and convert conversational data into actionable insights. Edgar's role focused on integrating SIP trunking across multiple phone numbers and client accounts, enabling large-scale, outbound and inbound call flows.

• Implemented trunks for several users;
• Implemented a solution for phone masking for certain clients;
• Created user-specific scripts that collected data for clients.

An app designed to track fitness progress. Since its launch, it has been a great success, with all 12 participants regularly using the app to track their fitness progress.

• Created the entire app in less than a week;
• Designed a user-friendly interface that made it easy for participants to track their progress and compete with their friends;
• Developed a robust backend system that ensured accurate fitness data tracking and maintained data privacy and security;
• Integrated various third-party APIs to provide additional features, such as data storage, GPS tracking in v2, workout logging, and sharing to Telegram;
• Continuously improved the app based on friends' feedback, adding new features and fixing bugs.

A company replaces traditional washing and drying machines in buildings with smart ones. The smart machines had an Arduino brain connected via MQTT to a central broker on a Raspberry Pi, which then connected to AWS.

• Created multiple apps that ran simultaneously on the Raspberry Pi, which enabled to control and monitor several smart machines at once;
• Optimized the code to reduce bandwidth consumption, allowing for smoother and more efficient communication between the machines, the Raspberry Pi, and AWS (modem ran on a 3g connection);
• Developed a remote updater that replaced a paid solution, allowing tumble to save costs while making sure that all machines were running on the latest software versions;
• Created an onboarding tool to deploy new devices on new premises easily;
• Developed software for the Raspberry Pi and made modifications to the Arduino;
• Implemented software for the Internet of Things (IoT);
• Improved the efficiency and functionality of hardware systems.

The goal was to reduce friction in loyalty and cashback programs between businesses, clients, and banks.

• Created a solid API for third-party services that allowed for smooth integration and communication between the platform and external services;
• Implemented a financial solution for the company by integrating Stripe, PayPal, and SPEI (Mexico's banking system);
• Developed a real-time data dashboard (React.js) that provided valuable insights into user behavior, transaction patterns, and overall platform performance;
• Created data transforming scripts in Python to handle large ingest of CSV files provided by the banks;
• Worked closely with this 3rd party team to improve the overall security of the project.

An initiative to provide relief areas for first responders and health workers working in COVID hospitals and staying for long shifts without going home.

• Set up the server on Digital Ocean Software development;
• Connected 3rd party APIs (Google maps, Firebase);
• Acted as support for all the team and people involved with pickup/delivery.

The company specializes in agile methodology in-person courses and seminars. The app required that the notifications and grading system be real-time, so a separate Node.js-Soketi server was implemented to use Websockets and achieve real-time communication between server and client.

• Created a fully-fledged solution, enabling the company to transition from in-person courses and seminars to a fully remote solution;
• Developed a web platform and mobile apps for iOS and Android using a single codebase in Flutter, resulting in an easier development process;
• Integrated 3rd party APIs and AWS services into the platform to provide a seamless user experience and high-quality content streaming (MUX, AWS wrapper);
• Led a team of one developer and one designer to deploy the platform and apps.