Hey, I am

Saman.

I love interacting with businesses to fetch real needs, turning them into technical handbook, and working towards clean, user-friendly software solutions. I help build systems, design interfaces, and ship projects that feel polished.

View Resume
Saman profile photo

Projects

A few builds I’m proud of — focused on real features, clean UX, and solid engineering.

Backstage Concert Companion Mobile Application

Concert-tracking app with a modern multi-tab experience, real-time event discovery, and a scalable Firebase account system.

KotlinAndroid StudioFirebase AuthFirestoreMVVMTicketmaster API
  • Developed a multi-tab UX for exploring events, connecting with friends, and managing interests using MVVM.
  • Implemented real-time concert search by integrating Ticketmaster’s API with location-aware content.
  • Built secure login + profile management using Firebase Authentication and Firestore.
  • Enabled saved events and synced cross-device data for a consistent user experience.
GitHubLive Demo

Ferry Reservation System

C++ system to manage ferry reservations and vehicle check-ins with persistent binary storage and SDLC-driven documentation.

C++Binary File I/OGitHubKanbanSDLCTesting
  • Designed and implemented a reservation + check-in workflow aligned with real-world constraints.
  • Produced requirements + design documentation (UI, architecture, detailed design) with explicit tradeoffs.
  • Implemented persistent data handling with binary file I/O; performed integration + unit testing (80%+ statement coverage).
  • Coordinated development via GitHub and Kanban to maintain transparency and consistency.
GitHubLive Demo

Squat Posture Classifier (Machine Learning)

KNN-based posture classifier using pose landmarks from video; reached ~80% accuracy through feature engineering and tuning.

PythonMediaPipescikit-learnKNNFeature Engineering
  • Built a KNN model to classify correct vs incorrect squat posture from video input.
  • Extracted pose landmarks using MediaPipe for reliable frame-level features.
  • Improved performance via feature engineering and model tuning to achieve ~80% accuracy.
  • Designed the pipeline for consistent real-time feature extraction from pose estimates.
GitHubLive Demo

Mobile Workout Tracking Application

Android fitness tracker with 3 workout modes (manual, GPS, and sensor-based inference) plus profiles and reliable local persistence.

KotlinAndroid StudioSQLiteGoogle Maps APISensors/ML
  • Built interactive flows to record, view, and manage workouts with a clean UI.
  • Integrated manual input, GPS tracking (Google Maps), and automatic activity inference using accelerometer signals.
  • Added a profile module with camera/gallery image selection, privacy preferences, and unit settings.
  • Improved reliability with SQLite storage + lifecycle-aware data handling for smooth persistence/retrieval.
GitHubLive Demo