Skip to content
View d2r3v's full-sized avatar
  • 19:41 (UTC -07:00)

Highlights

  • Pro

Block or report d2r3v

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
d2r3v/README.md

πŸ‘‹ Hi, I'm Dhruv Bhardwaj

Software Engineer Β· Systems & ML Β· UBC VEMS


πŸš€ What I Do

I build backend systems, ML pipelines, and interactive XR applications that are designed to work under real-world constraints.

My work focuses on:

  • Distributed systems & backend architecture
  • Applied machine learning (vision + NLP)
  • XR/AR interaction systems

I care about latency, reliability, and correctness over flashy demos.


πŸ”­ Selected Work

πŸ”Ή Exam Generation & Analysis System (Backend Lead)

  • Built a microservice-based backend (Node.js, tRPC, Prisma, PostgreSQL)
  • Processes 1000+ student submissions per exam cycle with automated grading
  • Designed snapshot-based grading system for reproducibility and auditability
  • Implemented RBAC + PostgreSQL RLS for multi-tenant security
  • Developed cheating detection and analytics pipelines

πŸ”Ή PillCare – Multi-Modal ML System

  • Built a vision + OCR pipeline (MobileNetV2 + CRNN) for pill identification
  • Designed modality fusion system to handle ambiguous inputs
  • Integrated OpenFDA APIs for real-time drug metadata

πŸ”Ή Companion – LLM-Based Support System

  • Designed a multi-stage LLM pipeline with routing, safety classification, and fallback logic
  • Built a persistent memory system for context-aware multi-turn interactions
  • Implemented risk-aware response handling for safe and reliable outputs

πŸ”Ή XR Systems @ UBC VEMS

  • Built real-time distributed systems for multi-user scientific visualization
  • Achieved <80 ms interaction latency using optimized WebSocket pipelines
  • Developed systems used in live classroom environments (100+ users)

🧠 Current Work

  • Developer Behavior Analysis (GitHub Data)
    Analyzing commits, PRs, and reviews to study collaboration patterns and detect anomalies

  • AR Interaction Research
    Exploring spatial anchoring and interaction stability in non-ideal real-world conditions


βš™οΈ Tech Stack

Languages: Java, Python, TypeScript, C#, SQL
Backend: Node.js, tRPC, REST APIs, Prisma, FastAPI, WebSockets
Systems: Distributed Systems, Microservices, Data Pipelines
Data: PostgreSQL, ETL, Query Optimization
Tools: Docker, CI/CD, Linux, Git


⚑ Fun Fact

Once spent an hour debugging a system issue that turned out to be a missing semicolon.
The system was fine. I was not.

Tech Stack

Primary

TypeScript C# Python JavaScript NodeJS Postgres tRPC Prisma

Systems & Tooling

Docker GitHub Actions

Data & ML

NumPy Pandas scikit-learn TensorFlow PyTorch OpenCV

UI / XR

React TailwindCSS Unity


Connect with Me

LinkedIn
Email

GitHub Stats






Dev Philosophy

"I tried to quantify burnout with data - turns out, I just built a mirror."
- Dhruv Bhardwaj

invisible counter

Pinned Loading

  1. PillCare PillCare Public

    Python

  2. reflect-ai reflect-ai Public

    TypeScript

  3. distributed-load-testing distributed-load-testing Public

    TypeScript 1

  4. TaskCollab/Task-Collab-BackEnd TaskCollab/Task-Collab-BackEnd Public

    Java

  5. TaskCollab/Task-Collab-FrontEnd TaskCollab/Task-Collab-FrontEnd Public

    JavaScript

  6. GroceSmart GroceSmart Public

    Java