https://code-blue.aiIdeated, designed, and built app which detects early signs of stroke and calls emergency services automatically. Though 1 in 4 people are expected to experience stroke in their lifetime, over 60% do not know the major signs, meaning Code Blue has the potential to save millions of lives- US Patent Office Provisional Patent Application accepted (Patent Pending status assigned)- One of the few teams selected for the I-Corps Lean Startup Bootcamp, conducting over 15… Show more https://code-blue.aiIdeated, designed, and built app which detects early signs of stroke and calls emergency services automatically. Though 1 in 4 people are expected to experience stroke in their lifetime, over 60% do not know the major signs, meaning Code Blue has the potential to save millions of lives- US Patent Office Provisional Patent Application accepted (Patent Pending status assigned)- One of the few teams selected for the I-Corps Lean Startup Bootcamp, conducting over 15 interviews in the span of a week- Demonstrated working prototype at the ASU+Global Silicon Valley Summit- Currently enhancing the app with additional functionalities such as transformation to camera plugin and balance detection Show less

Worked on context-backed chatbot and “co-pilot for democracy” products for users at the State Department’s Center for Analytics‬- Took ownership over 7 projects, building crucial products, tools, and functions that helped convert the pilot to full-time contracts‬‭- Expanded the testing of LLMs in workflow while simultaneously defining org-wide benchmarks on how to evaluate LLM performance‬ and safety (including red-teaming)‭- Developed solutions for context upload for chat agent… Show more Worked on context-backed chatbot and “co-pilot for democracy” products for users at the State Department’s Center for Analytics‬- Took ownership over 7 projects, building crucial products, tools, and functions that helped convert the pilot to full-time contracts‬‭- Expanded the testing of LLMs in workflow while simultaneously defining org-wide benchmarks on how to evaluate LLM performance‬ and safety (including red-teaming)‭- Developed solutions for context upload for chat agent using RAG, addressed 508 accessibility issues, tested and improved product‬ LLMs to generate greater accuracy, and shared tools across Palantir to benefit all commercial and government teams‬- Work with LLMs was demoed to Secretary of State Anthony Blinken and mentioned in his open address on the usage of AI in diplomacy‬‭Skills Leveraged‬‭ : NLP , RAG, red-teaming of LLMs, prompt‬‭ engineering, modeling, testing and benchmarking of LLMs, Python,‬ Javascript, Typescript, accessibility, data pipelining, object-oriented programming, ontology management, backend and frontend‬‭ Show less

Working at the intersection of cybersecurity and machine learning (secret clearance)- Designed and built time-series anomaly detection tool for log analysis using natural language processing (NLP) and hyperparameter optimization for data cleaning, pre-processing, clustering, and neural network classification- Demonstrated working neural network and clustering algorithms to Chief of AI & Data Science and department team members- Evaluated current implementation of Zero Trust… Show more Working at the intersection of cybersecurity and machine learning (secret clearance)- Designed and built time-series anomaly detection tool for log analysis using natural language processing (NLP) and hyperparameter optimization for data cleaning, pre-processing, clustering, and neural network classification- Demonstrated working neural network and clustering algorithms to Chief of AI & Data Science and department team members- Evaluated current implementation of Zero Trust cybersecurity architecture, culminating in whitepaper and plan for full incorporation and implementation by FY25- Briefed DCSA deputy director on log anomaly detection tool and Zero Trust gaps and implementation recommendations Show less

Working with University of California, Irvine, MIT, Harvard, and McGill University on the PyNIDM project, creating Python modules to analyze and understand the neuroimaging data model (NIDM)- Independently designed and developed research project using Python and machine learning to predict characteristics like hippocampus size based on certain factors such as gender and age.- Developed a clustering algorithm for NIDM documents providing the user four different types of clustering… Show more Working with University of California, Irvine, MIT, Harvard, and McGill University on the PyNIDM project, creating Python modules to analyze and understand the neuroimaging data model (NIDM)- Independently designed and developed research project using Python and machine learning to predict characteristics like hippocampus size based on certain factors such as gender and age.- Developed a clustering algorithm for NIDM documents providing the user four different types of clustering (agglomerative clustering, K-Means, Gaussian Mixture Modeling, and affinity propagation) to group data and show whether certain participants share a correlation.- Coded linear regression tool to find model coefficients, also including a number of features like co-variate capabilities, contrasting, and regularization- Worked with the Resource Description Framework (used to describe data frameworks on the semantic web) to create functions to transform the data into a usable format for machine learning.- Designed error correction and optimization functions.- Helped build graphing and visualization tools used to demonstrate the links between studies and participant data.- Research paper published in F1000 (https://doi.org/10.12688/f1000research.108008.1) Show less

Working at the startup SureStart to design and build an app helping Gen Z reach its future career goals - Wireframed the app on Figma, scripting out the onboarding process and dashboards for users- Individually built out a fully-functioning prototype on Bubble, creating features such as notification, email, and daily progress features- Spoke to investor to discuss how to raise funding

- Planned and taught lessons on Linux cybersecurity policies, servers, Cisco networking, and basic programming- Gained CTF experience with networking (routing, switching, firewall, packet tracing through WireShark) and understanding of HTTPS, DNS, TCP/IP architecture- Competed in the national CyberPatriot competition, specializing in Linux and server cybersecurity

Attended highly selective program to learn AI. Met speakers from MIT Media Lab, Google, Consumer Reports, Watson AI Team, and more- Learned about the entrepreneurial process, patenting, and leadership- Conceived & pitched an idea to create an app detecting facial asymmetry, one of the first signs of stroke- Created app (FAST – Facial Asymmetry Stroke Test) using AI and Python for competition in less than 2 weeks- Gave live demonstration of working application to MIT Media Lab… Show more Attended highly selective program to learn AI. Met speakers from MIT Media Lab, Google, Consumer Reports, Watson AI Team, and more- Learned about the entrepreneurial process, patenting, and leadership- Conceived & pitched an idea to create an app detecting facial asymmetry, one of the first signs of stroke- Created app (FAST – Facial Asymmetry Stroke Test) using AI and Python for competition in less than 2 weeks- Gave live demonstration of working application to MIT Media Lab Assistant Director, CSO of Consumer Reports, CEO of CySiv/Blockchain Founder, and CEO of Localized in final round- Won 1st place in the competition Show less