Jun 2022 – Aug 2022 · 2 mos · Cambridge, Massachusetts, United States · Remote

• Created multitask protein-ligand docking model and software, optimizing ligand enrichment and speeding up in-silico screening. Implemented semisupervised learning of protein/ligand representations to multitask learn binding pocket, pose, and affinity

Jan 2022 – Feb 2022 · 1 mo · Remote

• Worked on anomaly/risk detection within transaction graphs, and built a system to identify problematic clusters and future risk criteria

Sep 2021 – Jun 2023 · 1 yr 9 mos

• Building cognitively-inspired, open-source machine learning models and a suite of benchmark tests for supervised/RL research, based on a battery of cognitive assessments designed to test aspects of intelligence across species — aiming to establish similar benchmarks for artificial intelligence.

May 2021 – Nov 2022 · 1 yr 6 mos

• ◦ Developed no-code ML platform for biologists, from scratch – full-stack development. Built & optimized models across a variety of biological domains (proteomics, genomics, pathology, and more).
• ◦ Established infrastructure for data storage and distributed inference. Built website and model visualization tools.
• ◦ YC-backed. Currently working on deploying new models and data storage pipeline.

Sep 2020 – Sep 2021 · 1 yr · Cambridge, Massachusetts, United States

• Developed a computational pipeline to identify the clinical effectiveness of drug treatments from observational health record data, utilizing the identified signals to repurpose approved treatments for unmet medical needs.

Jul 2019 – Aug 2019 · 1 mo · MGH & BWH Center for Clinical Data Science, Boston, MA

• I worked at the MGH and BWH Center for Clinical Data Science to develop a semi-supervised NLP approach for fine-grained clinical report classification to optimize the process of cohort generation and model development. This was achieved through the creation of a deep unsupervised language model that automatically generated high-dimensional, contextual representations of medical reports. These report encodings, which contain complex clinical data, are then able to be used to train a classifier model with magnitudes fewer data and more accurately than was possible in the past. The encodings can be utilized as another data-dense modality for downstream patient analysis and predictive algorithms, while the semi-supervised model also optimizes cohort generation pipelines.

Jun 2017 – Jul 2017 · 1 mo

• I attended the LaunchX summer program at MIT during the summer of 2017. It was an intense month of intense work, where around 60 other students, handpicked from around the world, strived to build and grow a business. Influential entrepreneurs came in and taught about how we, as high schoolers could find a space to grow in the hustle of the business world. The entire team there is entirely supportive of young entrepreneurs and taught me that age does not directly correspond to the ability to create. I co-founded VocalEyes with two of my teammates during this time.

Sep 2016 – Sep 2019 · 3 yrs

• I am currently in charge of guiding companies at my school entrepreneurship club. I teach content to jumpstart students companies, as well as provide leadership and advice for students looking to apply for entrepreneurial ventures and endeavours in the area.

Jun 2017 – Sep 2020 · 3 yrs 3 mos · Greater Boston Area

• VocalEyes was cofounded at MIT Launch with the goal of appeasing the challenges that come to those with visual impairment. Having been personally affected by family members who suffer from macular degeneration, we seek to empower visually impaired people by building accessible technologies that are intuitive and useful. We continue to test our product at several institutions for the visually impaired across America, India and Egypt.

May 2017 – Jun 2023 · 6 yrs 1 mo · Allentown, Pennsylvania Area

• PlantumAI is an app that is able to detect, diagnose, and provide treatment options for plant diseases straight from your phone. If you're a farmer who needs help figuring what strain of disease, or what nutritional deficiency, is destroying your harvest, PlantumAI will be able to solve your problem instantly.

May 2017 – Aug 2021 · 4 yrs 3 mos · Allentown, Pennsylvania Area

• Lehigh Valley LaunchBox (LVLB) is a program of Invent Penn State, which aims to make the Lehigh Valley a hub for innovation and spur business startups across the Commonwealth.
• I was lucky enough to be accepted into the program, and they provide an enormous network of resources and people, which proved to be invaluable when starting PlantumAI.

Sep 2016 – May 2017 · 8 mos · Moravian Academy, Bethlehem, PA

• Open Agriculture Initiative (OpenAg) has a mission to create healthier, more engaging, and more inventive future food systems. It believes the precursor to a healthier and more sustainable food system will be the creation of an open-source ecosystem of food technologies that enable and promote transparency, networked experimentation, education, and hyper-local production. Working with a team of #nerdfarmers we built nation's first high school food computer (which we affectionately call as "Foam Computer") at Moravian Academy. I was involved in configuring the hardware and programming various sensors and actuators.

Aug 2015 – Jul 2016 · 11 mos