Medical AI · Robotics · Defense

SiddhantKhapra

I started building medical AI in high school. After my internship at Boeing, I got pretty obsessed with defense.

Now I want to build something where technology, medicine, and defense meet.

Some of my work

Right now

Connected health.

I’m researching portable diagnostic systems that can track small changes in a patient’s health and share updates across a mesh network.

  • Portable diagnostics
  • Continuous health monitoring
  • Mesh networks

Projects

Software and hardware

Boeing internship

Anti-submarine
warfare research

At Boeing, I used Systems Tool Kit and Python to model different maritime detection and strike configurations. I ran 54 simulations and compared which setups worked best in each scenario.

I presented my findings to more than 40 Boeing engineers and a U.S. House Representative. This internship is a big part of why I want to work in defense.

CycleAutomationStateMachine.javaJava
// Stop if a state takes too long
if (stateTimer.seconds() >= stateTimeout) {
    stop();
    return;
}

// Wait for the game piece
if (!isPieceDetected()) return;

advanceIntakeState();
ExtendLowerDetectComplete
Selected logic from my Columbia maker portfolioFull code

FIRST Tech Challenge

Robot cycle
automation

I wrote an asynchronous state machine to coordinate our robot’s intake and outtake mechanisms. The goal was to give the driver fewer things to manage during a match.

It checks a distance sensor before completing intake and uses timed steps to keep the mechanisms from interfering with each other. If a state takes more than 3.5 seconds, it stops and retracts the mechanism.

Read the code

Audio software

DesiDJ

My youth group didn’t have a DJ for a retreat, so I built one. You choose a Spotify playlist, and DesiDJ finds tracks that fit, matches their tempos, and mixes the transitions.

I also wrote a patent application specification describing the system, including text-based mood analysis, track selection, and the audio-processing pipeline.

AP Chemistry · LACSEF recognition

Solar hydrogen
experiment

I built an automated water-electrolysis setup for chemistry class and tested it with two power sources: a regulated 5V supply and a solar panel.

An Arduino-controlled pump moved the gas into a collection balloon so I could measure how much hydrogen the setup produced.

Steady 5V supply
~877 mL / hr
Solar panel
~88 mL / hr

If you want the details

The documents.