• Building AI models for risk assessment and triage in diseases like cancer, aimed at informing public health policy and reducing the burden on healthcare systems, advised by Dr. Danielle Bitterman under the Radiation Oncology department• Leading development of custom NLP methods, such as adapting the preference optimization algorithm ORPO to enhance the quality of synthetic medical data, improving the fine-tuning of medical AI models, and optimizing their applicability in clinical settings• Authoring a paper to NeurIPS' ML4Health symposium on how medical LLMs fine-tuned on synthetic patient data outperform those trained on gold-labeled data

• Working with Columbia CS Professor Luis Gravano on utilizing machine learning (ML) classifiers and large language models (LLMs) to identify potential cases of foodborne illness by analyzing natural language reviews on online review platforms• Collaborating with the NYC Department of Health and Mental Hygiene and LA County Department of Public Health to inform actionable public health policies

• Leading Reach4Help, a tech nonprofit connecting people in need to trusted volunteer help across 38+ countries• Managing a team building open-source software to coordinate logistics for humanitarian aid distribution in the most remote and underserved communities around the world alongside grassroots nonprofit organizations• Developed a need-matching web platform to coordinate food banks, clothing drives and vaccinations for 6K+ low-income communities• Developed a global map of 10K+ aid organizations, serving as the backbone for logistical coordination of volunteer help through the COVID-19pandemic, Mexican earthquake, California wildfires, Ukrainian invasion, Bangladesh floods, etc.• Raised $120K from Google for Nonprofits and another $106K in in-kind donations from Google, Slack, Algolia, etc.• Collaborated with the World Economic Forum’s Global Shapers in Europe to raise €170K+ for Ukraine medical aid relief• Collaborated with BRAC, JAAGO Foundation, Kandari and other Bangladeshi NGOs to serve 58K+ households affected by summer 2022 floods• Invited to give a talk at Google's HQ titled “The Roadmap to Resources: How Reach4Help Leads People in Need”

- Selected as a Future Leader by the Aspen Institute for my work with the Climate Reality Project on leading climate education projects in Bangladesh and now informing local climate policy around allocating funds in the US from the Biden administration’s Inflation Reduction Act (IRA)

• Spearheading local projects on immigrant youth empowerment in Queens and the greater NYC area. Still in its early stages, our main project is in partnership with local NGOs like 18by Vote and organizations like Google to extend resources to high school and college students in predominantly underserved and immigrant neighborhoods—to equip them with the skills to get more involved in tech and policy spaces• Have engaged in discussions on a future project for decarbonization via solar panel implementation in NYCHA housing• Elected as Curator (and formerly, Vice-Curator) for the NYC Queens Hub for 2023-2024• Coordinated SupportUkraineNow (SUN) projects for 1M+ people involving refugee aid, fundraising and mental health support as U.S. co-ambassador based in NYC (https://www.weforum.org/agenda/2022/03/supportukrainenow-lessons-from-young-people-humanitarian-response)• Participated in local community outreach initiatives for trash cleanup and wellness awareness in Queens, NYC with NGOs such as Woodside on the Move• Won a grant by the Climate Reality Project for leading a project to tackle lack of youth climate change engagement in Bangladesh• Invited to speak at the Climate Changemakers Digital Innovation for Youth Conference at the United Nations HQ in New York City• Selected as the youngest Global Shaper by the World Economic Forum for my work on Reach4Help

• Helped build self-supervised machine learning models for clinical usage in detecting stress patterns in electrocardiogram (ECG) data for early diagnosis and prevention of adverse fetal health outcomes like stillbirths and miscarriages• Developed a novel signal processing algorithm to separate out maternal and fetal ECG signals from abdominal ECG data• Published and presented at the 44th Int’l IEEE Engineering in Medicine & Biology Conference (EMBC) in Glasgow, Scotland as the lead author of a publication• Presented at the 55th Annual Meeting of the International Society for Developmental Psychobiology (ISDP) in San Diego, California as the lead author• Mentored incoming graduate students into the lab to onboard them onto the lab's biophysical signal analysis pipelines

• As member of the Bangladesh National COVID-19 Policy Dashboard Team of the Ministries of Health and IT, leading data analysis on coronavirus tests, cases, hospitalizations, deaths and mobility to inform actionable policy for lockdowns, public health communications and healthcare decisions• Statistically modeling COVID-19 spread to measure and respond to public health and socioeconomic effects of the pandemic to strike a balance between lives and livelihoods• Helped develop technologies for Bangladesh's self-reported disease surveillance system (before adequate testing capacity)• Co-authored several publications with epidemiologists from Harvard, Columbia, UC Berkeley, International Centre for Diarrhoeal Disease Research, Bangladesh, etc.• Coordinated with statisticians, researchers and data scientists using epidemiological indicators to develop decision-making tools for district administrators, policymakers and government officials up to the Prime Minister of Bangladesh

• Developed a pioneering “Right to Work” e-Commerce marketplace for Rohingya refugees with a team in Bangladesh to enable them to produce and sell products to consumers worldwide • Worked with the United Nations Development Program (UNDP), United Nations High Commissioner for Refugees (UNHCR), World Food Programme (WFP) and NGOs like ActionAid while stationed in the camps in Cox's Bazaar• Developed a mobile app being used in the camps to automate payment of hourly wages by scanning refugees’ UNHCR ID card

• Conducted computational biology research at Nobel Laureate Professor Joachim Frank's lab• Built 3D models of protein structures based on the various conformations of proteins in vitreous ice through cryo-electron microscopy for applications in drugs and treatments• Wrote a program for finding poor quality protein particles using micrograph defocus estimation values to detect and remove recalcitrant poor quality particles in our data

• Led a Stuyvesant HS team on a machine learning project under the guidance of a Google engineer to understand and emulate human emotions from voice intonations using LSTM RNNs and GANs to develop Digital Personal Assistants for administrative use, medical applications, speech disability, among others• Built two neural networks (one for Speech Audio Processing and Speech Text Processing) with ~80% accuracy in predicting emotion from an audio file (wav, mp3, etc) For more info: https://github.com/schowdhury8/C0FF33

• Facilitated biology and biochemistry research for 80+ students, linked club members to research lab principal investigators and competitions• Ran and published in the school-wide research magazine Sigma• Taught wet lab techniques and held workshops on how to write research papers• Had been an active member of the club since freshman year (2015)

• Selected by the Stand Up to Cancer Program in conjunction with the Weill Cornell Medical Center • Conducted my own research project and wrote a paper analyzing liquid biopsy data to find features between “clean” and “noisy” sites within the human genome for lung cancer patients• Helped develop recurrent AI models for early detection of cancer and tracking of response to treatments