Dyllan Hofflich Email & Phone Number

United States

o Working on power distribution PCB and RPi shield to control motors for a bipedal robot with VR integration. o Team also seeks to manufacture a mini-robot for youth education in collaboration with various companies, including SparkFun, DigiKey, and AutoDesk. o Added RFID sensing to mini-robot for converting physical blocks into Blockly code, used to.

Ithaca, New York, United States

o Provided academic support for students while teaching them how to use FGPAs with Verilog.o Collaborated with faculty to redesign lab aspects of the course to improve the course's learning outcomes.

o Designed and added asynchronous functionality to our graph database's python interface to greatly speed up data ingestion and graph traversal.o Spearheaded implementation and testing of our database’s Python interface for new API endpoints.

Ithaca, New York, United States

o Collaborated with Professor Donald Greenberg, alongside PhDs, Masters, and other undergraduate students to use redundancy in robot construction site scans to significantly reduce data size using a SLAM model to reconstruct the scene with a fraction of the stored images, achieving a 72.5% data reduction rate.o Led development of robotic quadruped.

New York, New York, United States

o Directed and primarily implemented DICOM SR & SEG export and input functionality for radiology annotations and reports for ease of creating medical machine learning models.o Co-led the auto-generation of medical infographics and short videos using LLMs to aid in patient understanding of their radiology reports.o Red-teamed our chat platform to prevent.

o Led implementation of TensorFlow support for graph ML platform across all available graph databases.o Applied critical analysis to existing TensorFlow graph ML approaches, aiding the company’s final decision in TensorFlow execution.o Reviewed existing codebase to diagnose and fix pre-existing problems in the platform.

o Wrote and conducted independent research on creating a classifier to classify frontal chest x-rays into AP and PA view, testing if it led to higher computer-aided diagnosis accuracy rates. (2021-2022)o Wrote and conducted independent research on utilizing the USE-QA (Universal Sentence Encoder-QA) model to answer COVID-19 questions with higher accuracy.