Peer to Peer Replicated and Fault Toleratant File System
Developed with Python sockets, and an algorithm was developed to replicate the file over nodes in the system to provide a failure resilient architecture
Hi! I'm Ambrose Tuscano, a Computer Science Graduate and a Software Engineer at Openprise.
Professionally, I am invested in Software/Web Development and Machine Learning - Data Science. But, my interests lie in a whole bunch of stuff. Name it and I would probably have read about in it.
I love trekking and playing Football. I am always up for trying new cuisines and desserts and meeting people.We can definitely connect over a cup of coffee and discuss about your company and projects, further how I can make a difference as a Full-time employee. Or we could just plan up a trek to Yosemite National Park. Just hit me up on any social media! Or, if you prefer email at ambrosetuscano@gmail.com
Major Life Events
I believe in using my skills to make life easy. What is programming if not automating a dead machine?
Events that shaped my career to date.
-
Machine Learning Intern at NXP Semiconductors
-
Graduate Assistant for CMSC 676 : Information Retrieval
-
Research Assistant for AI-SSGL Lab at UMBC researching on Fake News Detection. Profile at: https://ai-ssgl.umbc.edu/ambrose-tuscano/
-
Graduate Assistant for CMSC 455 : Numerical Computations
-
Vice Chair For UMBC ACM Team
-
Completed Masters in Computer Science from University of Maryland, Baltimore County
-
Graduated with a Bachelor of Engineering in Computer Science from University of Mumbai
Projects
Document Retreival System
Developed CNN layers from scratch to create a Deep Learning model to recognize guestures with accuracy of 76%
TV series Recommendation
Created Flask Api which used content based and collaborative methods to recommend users TV series, on a Android app with 86% accuracy.
How Fake Hate News affected politics
Analyzed news sources using BERT and other sentimental Analysis techniques to see how hate based news were used by fake sources to try and manipulate the 2020 US Election.
Diabetes Detection
Created a Heroku based service to detect Retinopathy by means of using Resnet, with an accuracy of 89% and used ResUNet to localize the detection.