01
AI & intelligent systems
Building agents that make decisions, learn from feedback, and solve problems with measurable strategies.
COMPUTER SCIENCE · MIT MANIPAL
Curious mind.
Useful things.
I’m Kaustubh, a computer science student who enjoys turning ideas into intelligent, useful software.
HEY, I'M KAUSTUBH ↗
BUILDING THINGS THAT THINK
01 / BUILDER
FOCUS
AI & SOFTWARE
AI & SOFTWARE
SCROLL TO EXPLORE
ARTIFICIAL INTELLIGENCE✳ALGORITHMS✳COMPUTER VISION✳SOFTWARE✳
ARTIFICIAL INTELLIGENCE✳ALGORITHMS✳COMPUTER VISION✳SOFTWARE✳
01 / A LITTLE ABOUT ME
I like understanding how things work, then building something with that knowledge.
I’m studying Computer Science at MIT Manipal. I enjoy experimenting with AI, breaking down algorithms, and making projects people can actually try. I’m most excited by the space where problem-solving meets product thinking.
More about me on LinkedIn ↗02 / WHAT I DO
02
Algorithms & problem solving
Turning complex problems into clear logic, useful experiments, and fast implementations.
03
Computer vision
Exploring how software can recognize real-world objects and turn visual input into useful data.
03 / SELECTED PROJECTS
My work.
Experiments, ideas, and tools I’ve been building.
WORDLE AI ARENALIVE DEMO ↗
SALET
SHARE
RECOMMENDED GUESSCRANEENTROPY · MINIMAX
01 / ALGORITHMS · INFORMATION THEORY
Wordle Solver AI
An interactive solver that compares entropy and minimax strategies, recommends guesses, and tracks remaining candidate answers.
A LITTLE BIT OF STRATEGY
242
4816
232644
8128256
AI AGENT REINFORCEMENT LEARNING
02 / REINFORCEMENT LEARNING
2048 AI
An agent trained to play 2048 using deep reinforcement learning. Built to experiment with learning, evaluation, and game strategy.
OBJECT DETECTED ✓
RECOGNIZE · RECOVER · REPEAT
03 / COMPUTER VISION · SUSTAINABILITY
Rebottled
An AI-powered reverse vending concept that recognizes bottles and supports traceable recycling and recovery data.
04 / TOOLS OF THE TRADE
My tech stack.
Tools I use to turn ideas into working prototypes.
PythonPyTorchGit
HTML / CSSJavaScriptOpenCV
C / C++Reinforcement LearningLinux
GitHubNumPyAlgorithms
Always learning. Always shipping small experiments.
05 / THE JOURNEY
Learning by
building.
NOW
Computer Science undergraduate
MIT Manipal · Exploring AI, algorithms, and software engineering.
PROJECTS
Building practical experiments
Wordle strategy solver, 2048 reinforcement learning agent, and bottle-recognition prototype.
NEXT
More open-source work
Improving my engineering fundamentals and building projects worth sharing.
06 / YOUR TURN
Have an idea?
Let’s build it.
I’m always up for talking about interesting projects, AI, and opportunities to learn by building.