Parking spaces. Lab equipment. Pedestrian movement. My projects turn each into a system that can be mapped, tracked, or measured. I’m a Computer Science and Engineering student working across full-stack applications, AWS infrastructure, computer vision, and parallel computing.

Projects  ·  Experience  ·  Currently  ·  Contact

Parking navigation and occupancy monitoring

A prototype for parking facilities with multiple entrances. It represents the facility as a graph, assigns an available slot for a chosen destination, and shows the route in a Flutter app. Computer vision supports parking-bay occupancy monitoring.

Result: [X] ms median allocation-and-route response time across 100 valid requests on the same demo floor plan.

Flutter   Node.js / Express   PostgreSQL   OpenCV

A clearer view of equipment across institutions

A web application for MAHE laboratory staff to manage inventory, export asset lists, request loans and transfers, and track approvals through role-based workflows and an audit trail.

Result: [X] assets searchable in the seeded MAHE demo inventory.

Next.js   TypeScript   PostgreSQL   Drizzle ORM

One movement model, five computing backends

A collaborative pedestrian simulation comparing sequential CPU, OpenMP, MPI, CUDA, and hybrid implementations. I contributed the OpenMP backend, shared metrics, hybrid integration, the live comparison interface, benchmarks, and report studies.

Result: [X]× CUDA movement-time speedup over sequential for 5,000 pedestrians and 100 steps.

The chart uses documented local medians from one RTX 3050 laptop. Backend timing scopes differ, so the ratios describe that benchmark rather than identical end-to-end run times. The simulation’s risk labels are model indicators and have not been calibrated to real-world injury outcomes.

C++   OpenMP   MPI   CUDA

AWS Cloud Solution Architect Intern · CloudCapt Technologies
January–April 2025

• Worked with EC2, S3, VPC, RDS, Application Load Balancer, IAM, and security groups.
• Deployed websites, configured domains and SSL certificates, and designed a three-tier cloud architecture.

• Strong: TypeScript, Next.js, PostgreSQL, AWS deployment.
• Working knowledge: Node.js/Express, Flutter, Python, Docker, Linux, OpenMP, OpenCV.
• Exploring: MPI, CUDA, PyTorch, TensorFlow.

• Building: DocMind, an evidence-grounded document Q&A system for larger PDF, DOCX, and TXT collections. I’m working from an architecture and documentation scaffold; the API, UI, retrieval pipeline, and evaluation harness are planned.
• Learning: hybrid retrieval with pgvector and full-text search, citation-grounded answers, and RAG evaluation with Recall@k and MRR.
• Looking for: Not currently looking.

• ParkEase — last pushed 5 Nov 2026
• Parallel Crowd-Crush Risk Simulation — last pushed 5 Nov 2026
• Lab Asset Management — last pushed 5 Nov 2026

_{Refreshed weekly from GitHub repository activity.}

Email · LinkedIn