01
Visual-Inertial SLAM
Tightly-coupled visual-inertial odometry using IMU pre-integration and stereo features for robust pose estimation in GPS-denied environments.
Dhruvil
Parikh
Robotics
Software
AI
I build systems that perceive, decide, and act. From autonomous drones with sensor fusion pipelines, to the perception models behind what they see, to the LLM agents that reason about what they’ve learned.
Scroll
§ 01About
I’m a robotics engineer working at the intersection of physical autonomy, computer vision, and applied AI. My work spans the stack: sensor fusion and SLAM on real hardware, the perception models behind what robots see, and the LLM systems that reason about what they’ve learned.
Today I write surgical robotics software at Globus Medical. On the side, I’m building Metrux AI, an honest way to interview engineers in the age of Claude Code and Cursor. Before that, I built the Robotics & AI department at Ideal Institute of Technology from zero: nine engineers, ten paying customers, $300K revenue in months. And before that, I co-founded a defense AI startup selected for DRDO’s Technology Development Fund at INR 10 crore (~$1.2M) before COVID forced us to sunset.
MS in Robotics from Northeastern University. BTech in Electronics and Communication Engineering from NIT Surat.
§ 02Experience
Apr 2026 · Present
Founder
Metrux AI · Boston, MA
- Teams everywhere are using Claude Code, Cursor, and Copilot, and nobody has an honest way to interview for it. metrux.ai is what I’m building to fix that.
- If you’re hiring, it’s a signal you can defend to your team. If you’re an engineer, it’s a way to find out where you actually stand, and to prove it.
- I’m building all of it solo: the challenges, the grader, the browser sandbox with Claude Code in it, the recruiter scorecards. AWS Lambda and Aurora under the hood.
- We don’t ask the model to score you. We ask it to find the evidence, and the score follows from what’s there.
- Measuring True eXecution with AI. (ME-TRU-X.)
Jul 2024 · Present
Software Engineer II, Robotics
Globus Medical · Boston, MA
- I work on Globus’ flagship robotics platform: ExcelsiusGPS for spine fusion, deformity correction, and cranial procedures like DBS, biopsies, and shunts, and the newer ExcelsiusFlex for total knee arthroplasty.
- Public company, $2.94B in 2025 revenue, and one of the few platforms in the market that pairs imaging, navigation, and robotics in a single workflow.
- It’s the most weight I’ve ever felt on a codebase. When the system runs, there is a patient on the table.
- From the patient side: James Buie (69) back exercising three days a week. Mona Schaechter (70) back in her garden. Both six months after robotic spine surgeries at MedStar Georgetown that used this platform.
Oct 2023 to Jul 2024
Assistant Director, Robotics & AI
Ideal Institute of Technology · New Jersey
- Built the Robotics & AI department from zero: recruited and led 9 engineers, set technical direction.
- Shipped custom delivery, cleaning, assistive, and humanoid robot software to 10+ companies.
- Generated $300K in revenue in months, projecting $1M+ annualized.
- Program still runs today and has grown into a broader AI center.
May 2022 to Dec 2022
Robotics Engineer
GreenSight · Boston, MA
- Drone software contributing to programs for NASA, DoD, and the US Army.
- U-Net semantic segmentation for agricultural drone imagery, IoU > 0.5 at 96.47% accuracy.
- Cut telemetry latency 83% by rewriting MAVLink to ride over LoRa for long-range UAV operations.
- Indoor localization to 3% error using ESP32 UWB DW3000 ranging.
Jan 2020 to Jul 2021
Co-Founder & CTO
AISafe Electronics Solutions · Gujarat, India
- Co-founded a defense AI startup incubated at PDEU; selected for DRDO’s Technology Development Fund at INR 10 crore (~$1.2M USD).
- Sunset and returned the funds before disbursement after COVID forced the call.
- Multi-camera Raspberry Pi rig with live streaming and pressure-triggered capture, 100% capture reliability in test.
- Deep-learning OCR for license-plate recognition, 99.81% character accuracy.
Jun 2020 to Sep 2020
Computer Vision Research Intern
SFR Medical · India
- Improved handwriting OCR with a custom CNN architecture, +10% over state-of-the-art baselines.
- Designed a wound-classification system using CNNs on low-resolution clinical imagery.
Nov 2018 to Dec 2019
Associate Product Manager
ABC Power Systems · India
- Drove market research and competitive analysis feeding a ten-year strategic vision.
- Co-authored the company’s Vision, Core Values, and B.H.A.G. with the executive team.
§ 03Selected Work
I.Robotics
02
Pothole Detector for Autonomous Vehicles
ROS-based perception pipeline that fuses RGB and depth to flag road hazards in real time on a moving platform.
03
Sensing & Navigation
GPS, IMU, and magnetometer fusion for vehicle state estimation, with Allan-variance noise characterization and trajectory reconstruction.
04
Mobile Robotics: Autonomous Navigation Stack
Perception, planning, and control stack for a differential-drive robot navigating unstructured terrain.
II.Computer Vision
05
Real-Time 2D Object Detection
Live object detection on streaming video with sub-frame latency, written in modern C++ on OpenCV.
06
Augmented Reality Engine
Camera-calibration-driven AR overlay that projects 3D content onto detected planar markers in live video.
07
Content-Based Image Retrieval
Feature-extraction and similarity-search system that ranks images by visual content rather than metadata.
08
Image Super-Resolution
Deep-learning model that upscales low-resolution imagery while preserving structural detail and edge fidelity.
III.AI & Language
09
AI PDF Reader
Retrieval-augmented question answering over PDFs, using LLM embeddings and semantic chunking to ground answers in source text.
10
AI Voice Assistant
Speech-to-text → LLM → text-to-speech loop that turns a language model into a hands-free conversational agent.
11
AI Chatbot
Conversational agent with context retention, built on the OpenAI chat completion API and a lightweight memory layer.
§ 04Education
Northeastern University
Master of Science · 2021 to 2023
Robotics · Boston, MA
Wearable Robotics. Advanced Machine Learning. Pattern Recognition & Computer Vision. Robot Sensing & Navigation. Robot Mechanics & Control. Mobile Robotics. Robotics Science & Systems.
NIT Surat
Bachelor of Technology · 2017 to 2021
Electronics & Communication Engineering · Surat, India
Engineering Mathematics. Microprocessor Peripheral & Interfacing. Embedded Systems. Advanced Processor Architecture. Digital Signal Processing.
§ 05Contact