Frontend
React
JavaScript
Material UI
HTML/CSS
Blazor
My Projects
I've worked on a variety of projects, from simple websites to complex web applications. Here are a few of my favorites.
Problem: I did not have a dedicated platform to showcase my work, skills, and experience in one place.
Approach: I built a personal site with portfolio case studies, contact flow, Neon Auth comments (Google/GitHub) with private image uploads, and a Gemini-powered AI assistant.
Results: Production site on Vercel with social login, visitor feedback, and MDX project write-ups.
Next.js
TypeScript
React
Vercel
Neon Postgres
Neon Auth
Vercel Blob
Google Gemini
Drizzle ORM
Cursor
Problem: Patients visit ERs blind to live wait times; capacity is opaque, so crowding and underuse happen side by side.
Approach: Full-stack app where facilities own their data; patients get a ranked, location-aware list sorted by average wait, with waitlists that update as they choose a site, plus history dashboards.
Results: Working prototype with dynamic re-ranking, real-time patient decisions, and admin visibility into visits.
React
Material UI
Node.js
Express.js
JavaScript
PostgreSQL
Problem: Accurate house-price models need large datasets moved through reproducible ML pipelines—manual notebooks alone do not scale.
Approach: Kaggle-based end-to-end pipeline on AWS: SageMaker notebooks, Lambda and Step Functions orchestration, EventBridge triggers, IAM-scoped access, S3 medallion layers, CloudWatch and RDS where needed, with Power BI for predictions, feature importance, and trends.
Results: Automated cloud-native pipeline from ingest through training and governed access, with stakeholder dashboards in Power BI.
Microsoft Power BI
Amazon SageMaker
AWS Lambda
AWS Step Functions
Amazon EventBridge
AWS IAM
Amazon S3
Amazon CloudWatch
Amazon RDS
Python
Problem: Face recognition is exposed to spoofing via prints, replays, and masks, weakening secure auth and monitoring.
Approach: Trained on 65,000+ images (~9 GB) of real vs. fake faces. Two-phase pipeline: YOLOv11 for live face localization, then a hybrid VGG16 + ResNet50 classifier for spoof detection—trained on Google Colab GPUs.
Results: 99.8% spoof-classification accuracy across diverse attack types.
TensorFlow
YOLOv11
VGG16
ResNet50
Pandas
NumPy
scikit-learn
Google Colab
Jupyter Notebook
JavaScript
GitHub
Problem: Newcomers to NL struggle to find reliable, centralized information on settlement, housing, health care, and community resources.
Approach: Backend API with FastAPI and MongoDB, Pydantic + Beanie for validation and ODM, Postman for API testing—structured to support a future newcomer-facing frontend.
Results: Actively developed API with validated settlement-resource endpoints and a scalable layout for the live product experience.
FastAPI
Python
MongoDB
Pydantic
Beanie
Postman
Previous projects
Earlier in my career, I designed and developed several enterprise applications:
Quality Management System
Oracle Forms Builder
Oracle Reports Builder
Oracle Database
SQL Server
QlikView
Warehouse Management
Oracle Forms Builder
Oracle Reports Builder
Oracle Database
SQL Server
QlikView
Purchasing
Oracle Forms Builder
Oracle Reports Builder
Oracle Database
SQL Server
QlikView
Automotive Auction
Oracle APEX
Oracle Database
Microsoft Power BI
All four ran as integrated modules within a large-scale ERP (Enterprise Resource Planning) system, sharing data across the enterprise.
Automotive Audit System
A platform for auctioning vehicles with minor defects: each car is inspected and documented, then listed in an online auction and sold to the highest bidder.
React
Redux
JavaScript
Node.js
Material UI
HTML
CSS
PostgreSQL
AWS RDS
AWS Lambda
AWS S3
Hackathons
Below is a hardware build from a local hackathon—quick prototypes, real constraints, and a lot of learning in a short window.
Hardware Hackathon — Just Do It!
Project: TennisBall Collector
Small-scale mobile robot built around a Raspberry Pi with motor control, power management, and sensing—prototype hardware to detect and drive toward a tennis ball, with demo branding for NL community tech events.
Skills
Technologies and tools I work with
Backend
Node.js
Express.js
FastAPI
Python
C#
ASP.NET
Databases
Oracle
SQL Server
PostgreSQL
MySQL
MongoDB
Cloud and data
IAM
SageMaker AI
Lambda
Step Functions
EventBridge
AWS Glue
Athena
Microsoft Power BI
Tools
Postman
Git / GitHub
Docker
Jupyter Notebook
Cursor
Claude
Certifications
Professional Development
Professional Skill Development Program
Memorial University
AI for Data Engineering
techNL
Artificial Intelligence Fundamentals
IBM
Education
Volunteering
Community involvement
Digital Seniors NL
Digital literacy for seniors
Memorial University
Student Volunteer Movement (SVM)
NL Eats
Community food support
Let's Talk Science
STEM outreach
Mineral Resources Review
Conferences
Iranian Moms of St. John's
Community group (IMNL)