Aaron's Portfolio

This is my first computer vision project. I worked on a team of 30 engineers as a junior engineer.  I am not authorized to publish source code. The stack we used was python of course, and we used google earth engine and ResUNet in two different models to do field delineation and crop classification of chili.    If you want to know how this project works I suggest watching this video.       I am not allowed to share the slides of the project either.   I was the data collection team manager and I managed the data collection fork of the repository for 8 weeks.  

One of the first project I completed for Omdena AI, was a redacting software that flags personal information from people.  The software uses a flask backend API, and uses Spacy.  I am not allowed to publish source code from this project, but here is the link to the website of the finished product. redactable-website Here is a raw output of what it looks like of data, as you can see it gives the page locations, and the relevant information that was redacted in the data output.  We did not build the front end, this was a long project and I am pleased to see it working. This project was my first end-to-end NLP (Natural Language Processing) project.  As you can see it is a quite intelligent project having both a backend and front-end team and is a workable AI product.