Oct 2026 – present
Technicaldle
A daily puzzle game for building probability intuition. Every region has one puzzle a day that everyone gets, so players can compare notes afterwards.
1 shared puzzle a day per region
TypeScript, deployed on Vercel

Oct 2026 – present
A daily puzzle game for building probability intuition. Every region has one puzzle a day that everyone gets, so players can compare notes afterwards.
1 shared puzzle a day per region
TypeScript, deployed on Vercel
Jul 2026
A small neural network library written from scratch in NumPy, with hand-derived backpropagation and no deep learning framework underneath.
0 deep learning frameworks used, NumPy only
Python, NumPy, Matplotlib
Apr 2026
A real-time 2D projectile tracker built on a discrete Kalman filter and a custom templated C++ matrix class.
90% lower latency than single-threaded multiplication
C++, multithreading, linear algebra, statistics
Jun – Jul 2025
A temporal GNN with a GRU encoder that predicts long-horizon realized volatility for 11 market sectors and beats GARCH(1,1).
2.4% better than a GARCH(1,1) baseline (3.37 → 3.29 loss)
PyTorch, temporal graph neural network, GRU
Mar – Apr 2025
A custom Gymnasium environment where an agent climbs a wall hold by hold, built for training RL algorithms like DQN.
4 discrete actions, grab, release, shift left, shift right
Python, Gymnasium, Matplotlib, reinforcement learning
Nov 2024 – Feb 2025
A PPO agent for 1v1 Tron, trained with behavior cloning and then a curriculum of stronger BFS and Voronoi opponents.
74.4% win rate against a random player
PyTorch, NumPy, Pygame, OpenCV, PPO
Jun – Jul 2024
A CNN autoencoder that compresses Mars surface images into 500-number encodings, then classifies landmarks from those encodings.
84.26% test accuracy classifying from the encodings
Python, CNN autoencoder, image augmentation

Long-form notes on reinforcement learning, filtering and volatility, with the math written out.
The first post is in progress. Subscribe by RSS to catch it.

Texas A&M University
B.S. Computer Science (Honors), minors in Statistics and Mathematics
Mathematical statistics and probability I & II, real analysis, linear algebra, multivariable calculus, machine learning, data structures and algorithms, discrete structures, software engineering, computer architecture.

Maroon Fund, Quant Researcher
Oct 2026 – present
Researching an inverse factor model that filters noisy features out of the broad tech market.
Aggie Data Science Club, Lead Workshops Officer
Oct 2024 – present
Leading deep learning workshops for 200+ members; technical games raised attendance 10%.
Looking for low-level, math, statistics and ML internships. Email is the fastest way to reach me.
arjun.mahableshwarkar@gmail.com