Prabhaav Pillai

Hi, I'm Prabhaav :)

Computer Science Graduate Student & Software Engineer

I’m an M.S. in Computer Science student at Michigan State University specializing in full-stack web applications, real-time systems, and computer vision. From architecting enterprise React and .NET solutions used by thousands of professionals to training and integrating real-time computer vision models for automotive systems, I focus on building clean, production-grade software that solves real-world problems.

Projects

Vision ASM: AI Blind-Spot Monitoring

Vision ASM: AI Blind-Spot Monitoring

Detects adjacent-lane vehicles in real time on embedded automotive hardware.

PythonC++YOLOONNXComputer Vision
Vision ASM: AI Blind-Spot MonitoringVision ASM: AI Blind-Spot MonitoringVision ASM: AI Blind-Spot Monitoring

A safety-critical, community-driven vision system built on the driver-facing wide-angle camera as a shipping production feature.

  • Appointed to Development Team with full repository write access to help steer project roadmap and backend architecture.
  • Authored and merged 1.7K+ lines of C++/Python into production, deploying a real-time blind-spot monitor to thousands of active vehicles.
  • Trained custom YOLO26n ONNX models on 27+ hours of crowdsourced driving video with semi-automated CAN-bus radar labeling.
  • Engineered edge performance on Snapdragon SoCs using zero-copy NV12 buffer slicing and CPU core isolation under 1.2x throttle overhead.
  • Drove community adoption (43K+ views, 187 upvotes on Reddit) and maintained the production lifecycle across bug fixes and dataset updates.
View Project
Picture-in-Picture Side Camera

Picture-in-Picture Side Camera

Side-window preview that appears only when you signal or a blind spot is detected.

PythonC++GLSLOpenGLComputer Vision
Picture-in-Picture Side Camera

A driver-assist overlay that surfaces the side view without the driver having to look away.

  • Appointed to the core StarPilot dev team with repo write access; authored and merged ~1K lines of C++/Python/GLSL to deploy the feature.
  • Wrote a custom GLSL shader (PR #83) that fuses Y/UV textures with a circular vignette mask, performing on-GPU NV12-to-RGB conversion for a smooth, cropped bubble preview.
  • Engineered zero-copy NV12 buffer slicing and reactive crop-mask configuration for per-vehicle camera calibration.
  • Built Galaxy web tooling with Flask REST APIs, a live frame snapshot endpoint, and a side-preview configuration UI with validation, plus unit tests for the device settings layout.
  • Wired the bubble to factory BSM, V-ASM, and turn-signal states so it only appears to provide additional lateral awareness.
View Project
StarPilot Galaxy: Mobile-First Remote Control Center

StarPilot Galaxy: Mobile-First Remote Control Center

The first progressive web app for StarPilot, now the default for every user.

VueJavaScriptPythonFlaskCSS
StarPilot Galaxy: Mobile-First Remote Control CenterStarPilot Galaxy: Mobile-First Remote Control CenterStarPilot Galaxy: Mobile-First Remote Control Center

Architected the Galaxy mobile-first Progressive Web App from scratch, authoring 22k+ lines of Vue/JS/CSS and Flask APIs to achieve 100% adoption and strong community acclaim.

  • Built Galaxy's settings, tools, navigation, tuning, models, Bluetooth/controllers, and device management into one unified frontend.
  • Reorganized every tuning control into clearly grouped sections and added a Theme Maker with a preset Theme Library for customizing the look of each style.
  • Added cookie-based mobile authentication and a fully responsive layout across phones and desktop.
  • Optimized for Firefox to remove lag on the heavily animated UI, and refactored the frontend into Vue native windows with consolidated, faster REST APIs.
View Project
UWM (United Wholesale Mortgage) IT Goals Dashboard

UWM (United Wholesale Mortgage) IT Goals Dashboard

Senior capstone built for a real enterprise client.

ReactASP.NET CoreC#Entity FrameworkAzure SQL
UWM (United Wholesale Mortgage) IT Goals DashboardUWM (United Wholesale Mortgage) IT Goals Dashboard

An interactive dashboard that turns goal data into clear, color-coded charts for quick trend analysis.

  • Track both current and historical goal data to compare performance across months and years.
  • Admins can add or edit yearly goals along with their metrics.
  • Built with a React front end, ASP.NET Core/C# REST API exposed via Swagger UI, Entity Framework Core, and Azure SQL Server.
View Project
Web Scraping & AI Content Recommendation Engine

Web Scraping & AI Content Recommendation Engine

An AI-powered recommendation engine built from scratch to scrape, process, and organize content.

PythonWeb ScrapingEmbeddingsTypeScript
Web Scraping & AI Content Recommendation EngineWeb Scraping & AI Content Recommendation Engine

Built for a CSE 881 data mining course project, it emphasizes a complete, from-scratch pipeline rather than an off-the-shelf model.

  • Scraped and parsed raw IMDb datasets (titles, crew, ratings) plus custom HTML scraping for supplemental metadata.
  • Built text embeddings from plot summaries to power cosine-similarity recommendations.
  • Shipped an interactive frontend for browsing and comparing recommendations.
View Project
Portfolio Website

Portfolio Website

The site you're looking at right now.

AstroTypeScriptCSSGitHub ActionsWeb Deployment/Domain Setup
Portfolio Website

A fast, effective Astro-powered portfolio that showcases projects, papers, and a resume. Designed for a custom domain with a card-based stack and expandable project details.

  • Built with Astro for static, prerendered output, deployed to GitHub Pages via GitHub Actions, utilizing a custom domain.
  • Project cards expand in place with a lightbox for images and PDF previews for papers.
View Project
PortVoy: Local Network Server Scanner

PortVoy: Local Network Server Scanner

Privacy-first Android utility with no tracking or ads.

JavaAndroidNetworking
PortVoy: Local Network Server Scanner

A lightweight, on-device tool for connecting to your own servers in a couple of taps.

  • Discover network interfaces (including hotspot connections) and scan for open ports.
  • One-tap connect to frequently used servers like FrogPilot or Portainer via a settings shortcut.
  • Fully open source and on-device only - everything stays on your phone.
View Project
Lon-Capa Physics Grade Calculator

Lon-Capa Physics Grade Calculator

Paste a semester of scores and watch the grade update live.

HTMLCSSJavaScript
Lon-Capa Physics Grade Calculator

Built for MSU physics courses that use the Lon-Capa platform, so students can see exactly where their grade stands.

  • Import lecture, homework, exam, and LearnSmart scores in one go.
  • Adjust weightage per category with live recalculation.
  • Automatic grade & GPA classification with a modern, intuitive UI.
View Project
StudyLink: Campus Social App

StudyLink: Campus Social App

Meet classmates, discover events, and join interest-based groups.

JavaAndroidFirebase
StudyLink: Campus Social App

A personal Android project for a CSE 476 course, built around Firebase for authentication and real-time data.

  • Browse and search student profiles, events, and groups from a single feed.
  • Secure sign-up and account creation with persisted user data.
  • Image saving/loading integrated via Firebase; polished profile and homescreen flows.
View Project
Phantom Traffic Jam Simulation Study

Phantom Traffic Jam Simulation Study

Studies how partial autonomy adoption affects traffic-flow stability.

JavaScriptPythonTraffic SimulationResearch
Phantom Traffic Jam Simulation StudyPhantom Traffic Jam Simulation Study

A collaborative study and multi-lane ring-road simulation testing whether Cooperative Adaptive Cruise Control (CACC) can act as a 'traffic shock absorber' in mixed fleets of human-driven and automated vehicles.

  • Modeled mixed-fleet traffic with IDM, CACC, and MOBIL in a ring-road simulation to reproduce phantom jams.
  • Varied CACC penetration from 0% to 100% across single- and double-lane scenarios, measuring flow stability and congestion.
  • Found meaningful stability gains from partial CACC adoption (peak improvement around 60%), informing V2X investment thresholds.
View Project

Papers

Adaptive LLM Load-Balanced Routing

Open

AV Penetration & Phantom Traffic Jams

Open

Learning vs. Heuristic Movie Recommendation

Open

Traffic Sign Physical Corruption

Open

Resume