Joined the final stage of a year-long project to develop an LLM-powered chatbot for Harvard Business Review, leveraging their extensive content archive to deliver accurate responses on leadership and management topics. Brought in to help finalize the project, I contributed significant, production-level components that are now integral to the latest version of the application.Highlights:• Developed a robust intent classification system to accurately identify user intent, enhancing the chatbot's contextual understanding and accuracy by handling certain user query intents differently.• Implemented a multilingual support system, enabling the application to serve a broader, diverse audience by recognizing and processing inputs in multiple languages.• Collaborated on evaluation metrics with HBP’s editorial and product teams to ensure brand-safe, high-quality output through structured testing and red-teaming exercises.• Developed key features of the chatbot that are now actively used by HBR subscribers, a portion of HBR.org’s nearly 10 million monthly users, to access high-quality, on-demand responses to leadership and management queries.This project culminated in a successful deployment, with my contributions forming a foundational part of the system in production today. HBP now plans to extend this GenAI infrastructure across new products, bolstered by institutional knowledge and adaptable system components.

During my last role, I had the opportunity to transition to a remote consulting role, which allowed me to realize how much I didn't know about all the other parts of the world outside of the northern New Jersey bubble I call home.After that contract ended I decided that now is as good a time as ever to travel while I am still young before getting into a new role. Here's a glimpse of where I went: - A month in Houston Texas- A month in Pennsylvania with family- Montreal for 2 weeks- Portugal for 3 weeks (Visiting the Azores was incredible!)- Costa Rica for 2 weeks with some of my closest friends- A month in Africa, 1 week in Morocco, and about 3 weeks in Ghana (Super humbling experience)- Paris for 2 weeks (got to watch an Olympic event!)- Spain for 2 weeksI am incredibly grateful for all the opportunities that have allowed me to explore what the rest of the world has to offer. Taking some time off has also allowed me to spend time helping my dad with his business, which he runs entirely on his own. While I have been on a break from working in machine learning, I have not been on a break from learning. With all the extra time outside of helping my dad, I have fully ramped up the self-studying and side projects. In this upcoming phase of my career, I'm actively striving to be more public with my work. One major step I've taken is to make virtually all of my code open-source. Additionally, I've launched a blog to enhance my writing skills and share insights into my learning and thoughts.As I reflect on my journey and the diverse experiences it has brought me, I am thrilled to embrace the next chapter of my career. AI and machine learning are continuously evolving reaching even more domains, and I eagerly anticipate the opportunity to find my place in the broad landscape. Wherever my path may lead next, I am excited to immerse myself fully, ready to embark on new adventures and make meaningful contributions to the exciting future of technology.

As a Machine Learning Engineer on a small, dynamic team, I played a key role in architecting and implementing a system of classifiers that integrated computer vision, NLP, and audio processing to enhance media pipeline capabilities for a Fortune 500 media giant.Highlights:• Automated 30% of the ad placement pipeline for new video content, resulting in streamlined processes and improved resource allocation.• Leveraged Python, Pandas, and AWS SageMaker to process terabytes of data, reducing data processing time by 16%.• Generated technical reports for stakeholders, translating complex data insights into actionable business decisions, improving strategic planning and operational efficiency.• Partnered closely with the client throughout each phase of the project, from gathering requirements and conducting research to managing data, designing and implementing solutions, and evaluating business outcomes.

As a Data Scientist on airport security contracts, I was tasked with various projects related to airport screening systems.Highlights:• Developed a Generative Adversarial Network (GAN) in PyTorch to expand high-quality training data by 50%, improving a system that screens 15 million pounds of cargo daily and reducing costs by 22%.• Enhanced an airport screening tool by refining feature embedding similarity with GAN models in PyTorch and NumPy, resulting in an 18% cost reduction.• Led development of a MongoDB x-ray image database, creating a prototype for synthetic training datasets with potential cost savings of over $0.5M by reducing manual dataset construction.• Supervised intellectual property disclosures from team research, highlighting innovative contributions to airport security technology.

As a Teaching Assistant, I provided students with one-on-one tutoring on Python programming concepts to improve understanding and mastery.Highlights:• Developed key aspects of course structure, ensuring a clear, consistent framework for learning.• Graded assignments and proctored exams, maintaining high standards of academic integrity and feedback.• Communicated daily with students via Canvas and email to address questions, provide support, and reinforce core concepts, leading to increased student engagement and success.

Implemented a DNA classification system, achieving nearly 50% reduction in error rates by leading the development of convolutional neural networks (CNNs) in Keras on imbalanced data sets by leveraging an HPC cluster using SlurmLed presentation showcasing research outcomes highlighting the most impactful findings for subject matter experts

Research assistant under the supervision of Dr. James Geller, Dr. Hai Phan, and Dr. Soon Ae Chun using deep learning models to develop a computational system to monitor and detect drug abuse in tweets collected across Twitter.Used the python scikit-learn library’s text feature extraction tools to develop a preliminary drug abuse detection model which reached an accuracy of 88% in determining whether tweets indicated signs of drug abuse or not.Used the python pandas library to clean the training data by analyzing all the data points to find inconsistencies, establish a standard formatting across various data sources, and validate the accuracy of the data upon completion. (More details about specifics in projects section below)

Efficiently cleaned and organized tables as needed while simultaneously managing other needs of the servers in a time effective manner.Monitored guests area and responded to requests or cleaning needs promptly.Maintained cleanliness in the dining area by cleaning tables, sweeping the floor, and removing the trash.

Performed food preparation duties such as preparing salads, appetizers, and cold dishes, portioning dishes and brewing coffee.Succeeded and thrived in the fast paced and high stress work environment.Built and maintained relationships with regular customers by listening to their needs and remembering their preferences.Responsible for keeping an organized inventory space in order to ensure that tasks were performed as efficiently as possible.

Trained new bussers on managing the floor in order to effectively handle busy shifts.Assisted waiters in the preparation and delivery of food in a timely manner.Respectfully interacted with customers and proficiently responded to their wants and needs immediately.Assisted the kitchen staff in preparing food for the upcoming shift.