started gait guardian

wright-application/stem-essay/main.tex

@@ -68,9 +68,10 @@ There was no time to test, hardly any to breathe, and before I knew it, we were
 finger hovering over the large play button. Time paused. I heard the buzzer and pressed play. Success.
 In two minutes and thirty seconds, we became league champions.
 
-But just seven months back, I didn't know what a variable was. I was fully into music, and programming was not even an afterthought.
+It was almost hard to believe that seven months back, I didn't know what a variable was.
+I was fully into music, and programming was not even an afterthought.
 It was mere coincidence that my neighbor (and good friend) decided to start a robotics team, and given the minimal investment, I
-joined. Like nearly all of my endeavors, my FTC learning started with a Google search. I was learning at a snail's pace, and it
+joined. Like nearly all of my endeavors, my FTC journey began with a Google search. I was learning at a snail's pace, and it
 had taken me two months to simply make a motor move. Soon, I was hooked. Like a sponge, I was absorbing everything I had to learn,
 and I had eventually taught myself enough Java to become a functional FTC programmer.
 
@@ -80,5 +81,14 @@ The joy I experienced wasn't just from our robot picking up and scoring pixels,
 was resulting in a tangible output that I could witness. It was that moment where I decided to pursue a STEM career. I was
 no longer just a high school student, I was a STEM student, and I was ready to help change the world.
 
+But that readiness was tested in September 2024. In a spur of ambitious insanity,
+I had committed to building a machine learning model to predict gait patterns in Parkinson's Disease for my sophomore-year
+Science Fair project. The problem: I had no clue how to. And so I learned. Python syntax, NumPy arrays, signal filtering,
+feature extraction, model architectures. I had entered a brand new domain, and each concept seemed to confuse me in a different way.
+After two months of painfully laborious learning, coding, and debugging, I was finally able to transform raw sensor data into a
+functional and accurate classification model. Somewhere between the first error message and the final 96\% accuracy, I
+had managed to absorb a new discipline by pushing myself into unfamiliar waters.
+
+
 
 \end{document}