• Project: To develop a clinically useful pediatric brain tumor segmentation model for detection and response assessment. Advised by Dr. Benjamin Kann in the Artificial Intelligence in Medicine (AIM) lab - https://aim.hms.harvard.edu/.• Role: I was the primary student investigator on this project. I formalized the initial problem, conducted necessary background research, curated the experimental design, trained numerous numerous 2D and 3D CNN models and evaluated these models, and achieved promising results. Currently preparing a publication submission on findings.• Approach: Challenges in this domain consist of limited data and large variations in size and location. Our approach included leveraging a larger corpus of adult data to perform transfer learning to the pediatric setting. In medical imaging, there oftentimes exists unlabeled datasets, and due to human-expert resource constraints these are unlikely to be annotated. In this project we incorporate unlabeled data into the training process to boost algorithm performance in a semi-supervised manner, showing that even unlabeled data can be valuable in model training.

Advised by: Dr. Adam Czajka and Dr. Kevin Bowyer• “Human-machine Pairing for Trustworthy AI,” Naval Surface Warfare Center, Crane Division (via Purdue University), June 2021 - Dec 2022 (lead PI: Dr. Adam Czajka); part of “Trusted AI: Trust Is Not Optional” (lead PI: Dr. Chris Sweet). - Role: Expanding our work on human-aided model training [3] to new domains such as deepfake detection and medical imaging(in-progress). Currently investigating the value of AI guidance in human examination.• “Synthetic Iris and Contact Lens Detection” (SICLD), FBI Biometric Center ofExcellence (via West Virginia University, USA), May 2020 - May 2021. - Role: Collecting and curating the largest known assembly of publicly available irisdatasets. Development of a way to incorporate human judgment into the training process of AI models. Released large journal paper assessing open-source methodologies on the collected dataset.• “Software Tool and Methodology for Enhancement of Unidentified DecedentSystems with Postmortem Automatic Iris Recognition,” National Institute of Justice, January 2019 - June 2021. - Role: Operated as the lead grad student on developing a new human interpretable solution in a pioneering study on using computer vision methods in iris forensics. Currently writing a new paper detailing the developments from this project. Learned new skills during this project such as hard-case instance segmentation using MaskRCNN and SegNet.• “Iris Liveness Detection Competition,” (LivDet-Iris, http://livdet.org). Co-organization of the iris liveness detection competition series (2013, 2015, 2017 and 2020). Co-organizers: Clarkson University (US), University of Notre Dame (US), Warsaw University of Technology (PL), Medical University of Warsaw (PL), West Virginia University (US), IIIT-Delhi (IN) and IDIAP (CH). Sponsors: CITeR (US) and NASK (PL). - Role: Designing the experimental pipeline and running competitors algorithms to attain scores.

Developed in Java on large scale publicly deployed applications to the highest coding standards. Learned about and employed the correct coding practices, approaches and methodologies. Worked in an Agile environment using Scrum. Located and fixed software bugs directly on the product and assured further quality of the software using manual testing and testing automation tools such as RCPTT.

Managed clients on a daily basis and worked closely with large teams to design and complete apps. Designed multiple websites, developed applications to client specifications and ran extensive testing to ensure quality. Led research efforts to determine market viability and compiled important information that would enable app creation. Graphic design for apps such as creating the UI/UX for applications. Developed apps for large scale companies from conception to completion. Successfully marketed products and learned how to effectively build a product's brand.

Taught freshman engineering students how to proficiently use AutoCAD. Communicated with professors every week to go over the assignments and organized what the key points of the assignment were and the best teaching approach specific to each problem.

• Surveying both buildings and lands with the latest GPS equipment.• Drafting with both AutoCAD and SolidWorks on an everyday basis.• Report writing for clients.• General office duties and telephone answering and logging.• Collaborated with clients constantly in order to design models of proposed structures on SolidWorks and AutoCAD.