Ethan Joseph
Machine Learning Engineer at OpenEvidence- Claim this Profile
Click to upgrade to our gold package
for the full feature experience.
Topline Score
Bio
Experience
-
OpenEvidence
-
United States
-
Software Development
-
1 - 100 Employee
-
Machine Learning Engineer
-
Jul 2023 - Present
Cambridge, Massachusetts, United States
-
-
-
Orbitax
-
United States
-
IT Services and IT Consulting
-
1 - 100 Employee
-
Machine Learning Engineer
-
May 2022 - Aug 2022
Remote Designed a recommender system for a digital marketplace used by hundreds of tax professionals. Built prototype for a natural language generation system utilizing GPT-3 to automatically generate country-by-country summaries for ESG reporting.
-
-
-
Substrate AI
-
Spain
-
IT Services and IT Consulting
-
1 - 100 Employee
-
Research Intern
-
Jan 2022 - May 2022
Remote Researched usage of neural attention mechanisms in reinforcement learning agents on visual environments, improved performance on certain Atari environments by over 200%. Wrote pipeline to improve training/evaluation efficiency by 5x, deployed parallel GCP compute instances to automatically train and evaluate different agent architecture configurations.
-
-
-
Rensselaer Polytechnic Institute
-
United States
-
Higher Education
-
700 & Above Employee
-
Undergraduate Researcher
-
Jan 2020 - Dec 2021
Troy, New York, United States Developed NLP powered web-scraper to extract vector-borne disease data from thousands of outbreak reports in collaboration with NASA Goddard researchers, published and presented at IEEE Big Data Workshop. Authored research paper on improving tabular data-to-text generation using transformer neural networks, outperformed previous SOTA by over 10% on certain metrics. Curated tasks and trained baseline models for novel dataset on common-sense inference.
-
-
-
Rensselaer Center for Open Source
-
Software Development
-
1 - 100 Employee
-
Project Lead
-
Jan 2020 - Dec 2021
Troy, New York, United States Led a team of over 15 developers from zero full-stack mobile or web dev experience to closed beta for smartrider, an open-source mobile/web app that consolidates all of RPI’s transportation services into a single location. Overcame a myriad of technical problems created by covid, including the complete restructuring of school transportation systems.
-
-
-
Software Engineering Institute | Carnegie Mellon University
-
United States
-
Software Development
-
700 & Above Employee
-
Systems Developer
-
Jun 2020 - Aug 2021
Remote Worked with Dr. Amit Vasudevan to develop the next-gen uberSpark toolkit for compositional verification of commodity system software. Developed hardware model abstraction to aid in automated program verification of hypervisor firmware. Converted open-source firmwares (TF-A/OpenSBI) to use uberSpark.
-
-
-
Deep North
-
United States
-
Software Development
-
1 - 100 Employee
-
Summer Research Intern
-
Jun 2018 - Aug 2018
San Francisco Bay Area Worked to develop and train facial recognition models for the WIDER Face Challenge 2018, scoring in the top 3 globally.
-
-
Education
-
Cornell Tech
Master of Engineering - MEng, Computer Science -
Rensselaer Polytechnic Institute
Bachelor of Science - BS, Computer Science -
San Francisco University High School