• Developed a customer facing web application that let users monitor products at e-commerce websites that included pricing, and product detail information using Python, Javascript, React, Kubernetes, Google Cloud, AWS Lamda and PostreSQL.

• Developed and managed a cluster of webscrapers/webcrawlers that obtained information from 15> e-commerce websites.

Leanglass

Jan 2019 – Present

Project description

Leanglass, personal project to create an online profile assessment of educational characteristics of a person for marketing and selling purposes:

• Developed an application based on Natural Language Processing that emulates a human's ability to assess a person's quality of writing with a >80% accuracy, outperforming most of published research benchmarks.

• Technologies used: Google BERT NLP (transfer learning), Tensorflow, Python, React, Node js, MongoDB, AWS, Docker, Docker Machine, Styled Components, Scikit-learn.

Deliverio is an application that solves the vehicle routing problem (VRP). Given a fleet, it is required to cover the demand of a set of delivery points efficiently. Starting at the same depot, each truck makes a round trip hopping on a subset of the delivery points. The goal is to minimize the overall distance. This is a generalization of the Travelling Salesman Problem (TSP) making it an NP-hard problem.

Technologies used:

• Google Maps API

• Docker

• React Native

• Node js

• Express

• Redis

• PostgresSQL