# Background

Growing up in South Bend, Indiana, I always enjoyed math. I initially focused on sports, but once I realized that I had more talent and fun with math, I leaned into it hard by self-studying MIT's undergraduate math/physics curricula. I also worked at Mathnasium on evenings and weekends and did some little projects on my own like finding/proving a formula for the partial fractions decomposition of $x^n/(x-a)^k.$ Two science teachers at my high school (Andrzejewski & Sisk) ran a research program that helped students connect with local university labs for science fair projects; I helped improve acoustic and optical data transmission in particle detectors and ended up making it to internationals.

## College

In 2014 I received a full ride to Notre Dame and became obsessed with modeling the human brain as a weighted graph. But aside from a highly-abstracted thought experiment that turned out to be mathematically solvable in a simple case, it seemed near-impossible to make progress from first principles! I also tried a top-down approach during an internship in Los Alamos that summer, attempting to create an emergent phenomenon of brain oscillations by implementing spiking neurons in a deep learning model. But training a spiking neural network turned out to be equally difficult and, beyond implementing the spiking neurons themselves, the project wasn't very fruitful.

During my sophomore year I interned as a data scientist at Aunalytics and took a variety of grad classes across pure and applied math. While I enjoyed working on projects for some applied classes like game theory, I lost interest in pure math and turned my focus towards working full-time at Aunalytics while finishing up my degree. My main projects centered on churn prediction and exploratory analyses in financial services / digital news / general subscription services, plus a side project evaluating the potential of topological data analysis (which turned out to be elegant in theory but not so useful for Aunalytics's practical needs).

At some point I realized that quantitative modeling was most valuable when paired with domain expertise, and although I was grateful to have broken into data science so early, I wasn't excited by any of the domains in which it was conventionally applied. I was only interested in tutoring and music production, and it was unclear how to leverage mathematical modeling in either domain. But I liked tutoring enough to keep working at Mathnasium on evenings and weekends throughout college, so I planned to throw myself fully into tutoring in a big city and see how far I could take it. LA was the obvious choice since it's where my girlfriend Sanjana was attending college.

After graduating from Notre Dame in 2018 I moved to LA and took on scattered work in math education: tutoring for many different agencies, developing content for several online learning platforms, teaching high school and weekend test prep, and even writing several textbooks for fun. I also enrolled in Georgia Tech's online MS CS, which seemed like a no-brainer for \$7k.

Gradually, all my teaching and content development work converged at the same organization, Math Academy, a highly accelerated 6-12th grade math program where 8th graders took AP Calculus BC and high schoolers studied a full undergradute math curriculum. Math Academy ran on top of educational software built by the founder, Jason Roberts, a serial entrepreneur who wanted to create and commercialize the ultimate online math learning system.

I got involved at the core of Math Academy's software during the summer of 2019. At that time, the software was a tool that Math Academy teachers used to create and grade assignments -- they would manually select problems from the database, students would complete the problems online for homework, and the software would automatically grade the assignment and keep track of a student's history. But Jason had his sights set on the holy grail of self-service learning, and he asked me to
1. develop an algorithm that would automatically create personalized assignments while leveraging effective learning techniques like spaced repetition and interleaving, and
2. help identify and resolve any issues that were preventing students from learning the material fully on their own.

By the end of summer the project was a success, upgrading the software from a manual assignment creation tool to a fully automated and personalized learning system that could effectively support independent learners (without any teacher).

We kept iterating on it, year after year:
• During the 2019-20 school year, we started out with a single independent learner, a student who was previously in Math Academy and had moved to another state. She learned AP Calculus BC using only the system (i.e. no external help) and got a 5 on the AP exam.
• During the 2020-21 school year and COVID-19 pandemic, the system proved to be significantly more effective than traditional remote instruction, and by spring 2021 nearly all of Math Academy's school classes were using it.
• During the 2021-22 school year, we opened the system up to commercial beta users and reached the point that the system was 4x as efficient as traditional in-person classes covering the same material. Seemingly impossible things started happening like some 6th graders progressing all the way from prealgebra to AP Calculus BC in a single year.
• During the 2022-23 school year, we put up a website and began to scale.

### Eurisko

From 2020-23 I also taught Math Academy school classes and (in collaboration with Jason) developed Eurisko, a special applied math/CS sequence within Math Academy's school program, with the intent of pushing Math Academy's best students to the full extent of their abilities. The Eurisko courses were presented at a level of intensity comparable to those offered at elite technical universities. Students were not allowed to use external libraries – instead, they wrote everything from scratch. They started by implementing a variety of algorithms and data structures including matrix and graph classes, after which they learned supervised machine learning and implemented advanced models such as neural networks. The program culminated in reimplementing research papers in artificial intelligence (David Fogel's Blondie24 Checkers Player) and working collaboratively on a large-scale project (implementing an extremely complex board game called Space Empires).

I'm moving to Boston in June 2023 (working remotely). Sanjana is doing her PhD there, and we've endured long-distance since early 2020.

## Acknowledgements

Thank you to the following people who have helped me find my way:
• 1996 - present | My parents - for always encouraging and wishing the best for me
• 2017 - present | Sanjana Kulkarni - for being my other half
• 2019 - present | Jason Roberts - for helping me find my niche after college and providing a constant stream of challenge and mentorship
• 2015-17 | Dave Cieslak - for hiring me as a data scientist during college and sending interesting projects my way
• 2013-17 | Cari Ingram - for hiring me as a Mathnasium instructor during high school and letting me have so much fun goofing around with the kids
• 2012-14 | Ken Andrzejewski and Douglas Sisk - for teaching me how to put myself out there, find research opportunities, and compete in science fairs during high school
• 2012-13 | Ilan Levine - for advising my first high school physics research project