Dhruv Shah
I build mini self-driving cars, autonomous drones, facial recognition systems, mobile apps, websites and in my free time study robotics and management at the Jerome Fisher M&T program at UPenn. I'm a goal-oriented performer who strives for perfection and doesn't tolerate anything less. I've accumulated over 100,000+ downloads total on iOS app store over all my apps, have five hackathon victories, and am a Platinum division contestant in the national USACO Competitive Algorithmic Programming contest. After being invited to AngelHack's exclusive 3-month startup accelerator program, I got the wonderful opportunity to grow and pitch a startup along with two partners in Silicon Valley to top investors from YC, 500 Startups, AngelPad, TechCrunch, and more at the annual Global Demo Day. Under the mentoring of the experienced AngelHack team, I learned a lot about startup growth and designing for the customer, and I'm ready to apply that knowledge, hard work, and channel the drive for perfection into a beautiful product. I have extensive experience in applications of machine and deep learning - especially in the field of robotics, mobile & web development, UI/UX design, and backend/server architecture.
Connect with me! I'd love to hear your story!
Education
Saratoga High School
GPA: 4.00
Honors & Awards
- Sense360 Hackathon 1st place overall (remote)
- Angel Hacks (Silicon Valley) 1st place overall (out of 100 teams) - 2017
- HACKcelerator 2017 bach. Presented at GDD - November 2017
- LA Hacks winner of $2000 grant from 1517 fund (out of 300 teams) - April 2017
- LA Hacks Top 12 Hack
- USACO Platinum Contestant
- AMC Honor Roll
- Robotics - FTC Control Award (3x) - 2017
Courses & Certifications
- Udacity - Introduction to Artificial Intelligence
- Coursera/Stanford - Machine Learning
- Udacity - Self-Driving Car Engineer Nanodegree
Projects
I built my first iOS App in sixth grade. I released AnswerThis, a trivia game that utilized SQL and match-making backend technologies, in 2014. Before it was taken down because it was outdated, it garnered just over 50,000 downloads.
A spaceship 2D game developed with the team at Dali Labs.
A home automation control system, inspired by Marvel's Iron Man, that has language understanding capabilities and learns my habits and preferences, using them with full control over my pool monitoring/cleaning system, front door, my home lights, the security system, thermal system, and my garage door.
Hackathon project developed at Capital One 2015 SmallBizDev Hackathon SF: Reddit + Yelp combined for small businesses.
Waypoint connects businesses in travel, restaurant, and other industries to their users with immersive augmented reality (AR) experiences created by enabling interaction of the user’s phone with their physical surroundings via protected or public augmented-reality "geo-pins". Hackathon winning project-turned-to-startup through AngelHack HACKcelerator program.
ScheduleMe is a utility bot for small businesses to plan out and delegate their already-packed schedules, utilizing Microsoft NLP Libraries and the Microsoft Bot Framework to bring 21st century technology to small business owners. LAHacks 2017 Top 12 Hack & Peter Thiel 1517 Grant Winner.
Took an old RC car, rewired the motors to a Raspberry Pi with a camera with the eventual goal of Level 3-4 self-driving autonomy in self-constructed roads and tracks. Click to see this project's full journey!
The result of an exquisite fusion of hardware and software ingenuity, the Shah Autonomous Drone project was born out of boredom and the love for drones and autonomy. The goal? To be able to gesture-control a drone in video recording mode, be followed by a drone, and be able to run flight missions.
Udacity Advanced Self-Driving Car Engineer Nanodegree
Vehicle Detection and Lane Lines: This project involved taking prerecorded video input of dashboard camera footage of driving and running classifiers for vehicle detection and tracking, as well as lane line finding and tracking with polynomial fits.
Behavioral Cloning: This project involved using Udacity's vehicle simulator to record footage of me keyboard-driving a model car around a track. Using this footage paired with the NVIDIA model architecture, I trained a neural network to steer the car around the track.
Traffic Sign Classifier: The Traffic Sign Classification project implemented a modified LeNet Neural Network architecture that classified traffic signs.
Unscented/Extended Kalman Filters: Implemented Kalman Filters to smooth and predict actual 2D vehicle location from "radar" and "LIDAR" sensors.
Kidnapped Vehicle: Implemented a particle filter (SMC) to smooth and estimate actual 2D vehicle location from noisy sensor location data.
Model Predictive Controller: Implemented a model predictive controller (MPC) as a path-following motion control method that controls actuators (throttle, break, steer) on a model car to make it follow a trajectory of points around a track.
PID Control: Implemented a PID controller as a path-following motion control method that controls actuators (throttle, break, steer) on a model car to make it follow a trajectory of points around a track.
Skills
- Machine Learning/Artificial Intelligence
- Back-End Web Development
- Front-End Mobile & Web Development
- User Interface/Experience Design
- Computer Vision
- Robotics
- Hardworking
- Passionate
- Works well with others
- Results-oriented