Mauricio Pisabarro

A microservice architectured system, including services in Node.js with TypeScript, and IBM's Watson Assistant as the machine learning powerhouse. The system would parse social networks and process the data through pipelines so that it could identify potential customers to the company I was working at the time, based on the content of posts. In parallel, the system was capable of learning in pseudo-real time through manual tagging of examples and counter examples.

I designed, implemented and successfully deployed the whole system. The system allowed the company to get several leads on customers, but it found its biggest success as a showcase for the company's capabilities.

Developed an API similar to Booking, Kayak, or Despegar, capable of searching and booking flights, hotels, transportation services, restaurants, activities, and many other categories in the future.

Working on the redesign and re-architecture of the back end, I aimed to improve the performance needed to support the scalability issues caused by getting Google as a partner, showing results on their Travel API and allowing for future extensibility.

I got to improve latency by 10x by redesigning the current solution, and further 10x improvement is expected once the new cache architecture is implemented.

I could also significantly improve extensibility and modifiability, allowing for the faster onboarding of new developers, as well as achieving ease of addition of new search categories and data providers.

An ad-hoc solution for a marketing company can consume offline events such as purchases and detect whether those could be attributed to customers having seen ads. With such a tool, our client could administrate and analyze the performance of ad campaigns for their clients.

The whole system was built on top of AWS to be scalable, with several services comprising the back end, primarily written in Java with Spring Boot, and the front end as a single-page application written in React with TypeScript.

An Ionic-based app consisting of a marketplace through which one can make slates or packs of plants and building materials for a third party. This involved working closely with designers to achieve a rich mobile UI and UX and transforming an old SQL database (10 GB) to migrate data to be used on this new project.

I worked on the design and implementation of the full stack mainly on my own, except for the UI design and UX, which were solved mainly by specialized designers.

Monetization was a hard aspect of the project as both Android and iOS have different requirements on how payments are made, and we aimed to build a generic app for both.