Product Manager at Microsoft AI, Building systems that connect LLMs to human preference.
A brief overview of my career journey.
I'm passionate about using AI research in building tools for underserved industries - I've built Operator for folks who are disabled, Deep Research for Nonprofits, and ChatGPT but ranking for those unable to express preferences with words. I love chatting with people to iterate fast - I chose product after reading the Design Sprint and loving the idea of quickly falsifying hypotheses, so 4 months after I joined Microsoft AI I became the first PM in 2 years in the org to directly interview users, and founded a cross-org 10+ team of PMs and researchers that led to 40+ new research studies. I've always been driven by curiosity and a desire to solve meaningful problems, which has taken me on an unconventional path—from studying religion in a monastery to performing stand-up comedy, shadowing surgeons, and building tools that help millions of users.
Beyond tech, I've tried to explore life deeply: finessing my way into the Library of Congress with an email receipt to research PM methods when it was closed, building ML models to predict satisfaction based on college major choice, and coding tools for students and nonprofits.
I explored politics (Congressional internship), medicine (shadowing a surgeon), law (interning), and consulting before reading The Sprint Book & Algorithms to Live By, and loving the idea of building products people love and falsifying hypotheses. I then ran design sprints at Deloitte, built an ML model for my thesis predicting satisfaction with major choice, and built projects helping students (Wyrrutgers, Rustewed, TheLX, PocketCoach) & nonprofits (Care Somalia, Career Office). At Microsoft AI, I sharpened my experimentation and data-driven decision-making skills while building AI-driven products at scale. Now, I'm building AI agents to bridge the gap between research and real-world impact.
Featured projects that are public right now- across AI, web development, and mobile.
Developed a novel ranking-based LLM interaction model, enabling structured decision-making using AI. Evaluated ranking-based vs. chat-based AI usability.
🚀 Explored human preference modeling & AI alignment.A chrome extension that tells you how many calories are in your favorite menu items online using Google Review photos
🚀 Explored AI Agent grounding through navigationComparison on 5 use cases on accuracy and speed
🚀 Explored accuracy/speed tradeoff of AI agentsGenerated audio captchas and explored if GPT4o could solve it if we let it use a transcription tool
🚀 Explored intersection of AI agents and accessibilityIOS App that plays a motivational message when runners fall below their pace
🚀 Voted 1st in most creative project by CS Faculty at Rutgers in Student showcaseAn improved Ratemyprofessor for Rutgers students to choose professors
🚀 Reached 700 students in 3 days, led as PMLets you choose a prompt that you'll get a new image of every day
🚀 Explored image generation models and have photos that inspire myself when I start my dayWebsite for Rutgers students to share recipes using dining hall ingredients
🚀 Reached a few hundred students, learned React and building with user-generated contentLets you run ranked-style polling in your discord channels
🚀 Won 1st place at Rutgers Hackathon, learned JavacordChrome extension that helps you make a schedule by using an optimization algorithm
🚀 Explored building chrome extensionsFun script that presses spacebar a million times to scroll down your history to a specific date
🚀 Solved a problem I hadForum for upperclassmen to give advice to underclassmen at Rutgers
🚀 Tried no-code builders like Bubble.io, learned importance of having content to solve cold-start problems of feature launches
I'm looking to connect with:
✅ AI labs, researchers, and builders working on research that can be applied to real-world problems
I'm focused on AI agents, human-AI collaboration, and scaling AI in real-world settings. Let's chat!
📩 DM me or reach out below.