←Dhruv Chavan

Dhruv Chavan

Data Science & AI · Data Analytics at Ohio State

I build ML pipelines and LLM tooling, from spatial analysis of 1.2M+ data points in cancer research to GenAI apps at BMW Financial Services.

Experience

Summer Analyst Intern

Jun – Jul 2026

JPMorgan Chase & Co.

  • Automated vendor performance reporting in Snowflake, streamlining analytics delivery to senior executives
  • Conducted risk assessments on new initiatives to detect fraud patterns and support compliance reporting
  • Built authentication risk dashboards to surface fraud insights and monitor performance within tolerance

Snowflake · Risk Analytics · Dashboards

Data and AI Engineering Intern

Sep – Dec 2025

BMW Financial Services

  • Built MCP server tools connecting internal LLMs to PostgreSQL, improving data retrieval efficiency by 45%
  • Developed 3 GenAI apps spanning fraud detection, business intelligence, and call center automation
  • Launched an AI governance platform enabling 95+ employees to submit model requests for compliance review

Python · PostgreSQL · Docker · LLMs · MCP

Controls and Navigation Research Lead

Jan 2025 – Present

OSU Dept. of Mechanical & Aerospace Engineering — Dr. Mrinal Kumar

  • Architecting autonomous wildfire drones from scratch, calibrating Pixhawk with fused RGB/IR sensor streams
  • Deploying CNN segmentation on Jetson Nano to classify fire-risk terrain at 30 ms and 92% accuracy
  • Validating flight and mission logic in pymavlink SITL simulation, cutting field-test iteration cycles
  • Leading 5 students building ground robots that relay data to drones and ignite controlled test fires
  • Building multi-node ROS2 stack for LiDAR, AI camera and smoke sensors; wrote custom camera drivers
  • Achieving 2 cm aerial and 5 cm ground accuracy via Marvelmind indoor GPS with Nav2 path planning

Python · MATLAB · C++ · ROS2 · Kalman Filter · CAD

Undergraduate Student Researcher

Jul 2024 – Present

OSU Wexner Medical Center — Assistive Technology

  • Analyzing 4,000+ patient data points; findings presented to clinicians led to a 25% improvement in outcomes
  • Applying hypothesis testing to identify demographic correlations and visualize them for clinicians
  • Published as second author at RESNA; presented at 2025 RehabWeek to 2,000+ attendees

Python · Pandas · Power BI · MySQL · Matplotlib

Undergraduate Computational Biology Researcher

Feb 2024 – Jun 2025

Nationwide Children's Hospital — Immune Response Modeling

  • Analyzed 1.2M+ spatial data points from 30+ brain tumors, revealing cell clustering linked to ICI response
  • Developed advanced visualizations to analyze brain cancer cell interactions with immune cells
  • Second author on peer-reviewed CSBJ paper analyzing tumor structures across cell types
  • Built permutation-testing pipeline running 10,000+ randomized trials per slide to test significance

Python · R · Pandas · Spatial Analysis · Statistical Modeling

Projects

Autonomous Lego Technic Bugatti

May 2026
  • 80% lane-following accuracy with an adaptive-thresholding CV pipeline at 30 fps, publishing steering at 20 Hz
  • Cut path deviation by 75% by fusing IMU and camera with an Extended Kalman Filter

Python · OpenCV · ROS2 · EKF · C++

President's Buckeye Accelerator Program

Jan – May 2025

Keenan Center for Entrepreneurship

  • Developed a YOLO-based computer vision prototype assisting 5 test users with facial recognition deficits
  • Conducted user research through surveys and video calls; 95% of respondents validated the idea
  • Selected through competitive workshops focused on market strategy, design and entrepreneurship

Python · YOLO · OpenCV · User Research

Lymphedema Scanner App

1st place, OSU Appathon
  • iOS app detecting lymphedema with 95% accuracy, built with ARKit and SceneKit using vector grids for facial volume capture
  • Firebase for authentication and secure scan storage

SwiftUI · ARKit · SceneKit · Firebase

Animal Image Classification CNN

Jun 2024
  • CNN in TensorFlow/Keras reaching 98.2% accuracy across 5 animal classes, with data augmentation for robustness

Python · TensorFlow · Keras

Personal projects

Digital Terrarium

Rust · WASM · 2026

An artificial-life research platform where creatures with evolving neural-network brains forage, hunt, share energy, and reproduce in a seeded ecosystem. A Rust simulation engine powers both headless experiments and an explainable Three.js browser demo through WebAssembly. The demo includes a live story feed, creature and brain inspectors, and a measured energy budget showing food inputs, metabolic costs, transfer losses, and reserve changes. Research tools include validated parameter sweeps, versioned run records, genome exports, and exact checkpoint/resume within a compatible build. Controlled tests cover environmental and inheritance mechanisms. Open-ended evolution and cognitive gains remain research questions; earlier results affected by energy-accounting bugs are explicitly withdrawn. Passing builds redeploy the linked demo automatically; its header identifies the deployed commit.

Rust · Neuroevolution · WebAssembly · Three.js · rayon · Python · Reproducible Experiments

Live 3D demo → GitHub → Research progress →

Notch

macOS · Swift

A free, open-source study companion that lives in the MacBook notch. Hover over it and it expands into a Pomodoro timer, assignment tracker, calendar, now-playing controls, file shelf with PDF conversion, clipboard history and notes. Runs fully local with no dependencies.

Swift · SwiftUI · AppKit · EventKit

View on GitHub →

BlockCraft DS

Nintendo DS · C

A block-building survival game for the Nintendo DS and DSi, written from scratch in C. Procedural worlds with biomes and caves, around 100 crafting recipes, mobs, day/night cycle, flowing water and lava, touch-screen inventory and SD card saves. All art, sound and music are generated from code.

BlockCraft DS scenes: plains, sunset, a lava cave and a lake

C · libnds · Procedural generation

View on GitHub →

Publications

Skills

Languages
Python, C++, C, MATLAB, R, SQL
ML/Perception
PyTorch, YOLO, OpenCV, scikit-learn, Pandas, NumPy
Robotics
ROS2, Nav2, Pixhawk/PX4, pymavlink, Kalman filtering, sensor fusion
Hardware
Jetson Nano, Raspberry Pi, Arduino, BNO055 IMU, LiDAR, Marvelmind
Tools
Git/GitHub, CMake, Docker, Fusion 360

Leadership

Calculus Tutor

Jan – Aug 2025

Student-Athlete Support Services Office at OSU

  • Tutored 3 student-athletes weekly (~10 hrs/week) in Calculus I & II around tight travel schedules
  • Developed personalized lesson plans; received consistently positive biweekly evaluations

Mount Scholars Leadership Society

Aug 2023 – Present
  • Led Team Security initiatives and developed leadership through weekly training, mentorship and retreats
  • Completed 75+ hours mentoring athletes with Down syndrome through weekly tennis sessions and events
  • Mentoring underclassmen as a designated "Mega," providing guidance on academics and early college transition

Contact

Open to data science, ML engineering and AI research roles.