Ivan Tran

Completed 2 years of coursework covering software engineering, data structures, algorithms, and system design.

Multi-tenant franchise management platform inspired by inefficiencies observed at Walker's Donuts. Targets businesses that rely on repetitive task execution and need consistency across locations — restaurants, cafés, retail chains. Roadmap includes a knowledge-sharing hub for franchise communities. Deployed and live on Vercel.

• Architected multi-tenant RBAC with franchise hierarchy — 14 Prisma models, 6 permission types, organization-scoped data isolation on PostgreSQL (Supabase)
• Timetable system with weekly/daily views, role-based task filtering, and status tracking (TODO → IN_PROGRESS → DONE)
• Template engine for reusable weekly schedules — create once, apply to any date range with frozen task snapshots
• Invite system with accept/decline flow, seen/unread tracking, and real-time notification panel
• Auth.js v5 with Google OAuth, JWT sessions, httpOnly cookies; route protection via Edge middleware
• Service Layer architecture with discriminated unions for type-safe error handling
• Conducted structured V1 smoke test — 87 test cases across 18 categories with automated RBAC route testing; CodeRabbit AI review on every PR
• Live: friend-chise.vercel.app | GitHub: github.com/IvanTran-2001/FriendChise

Multi-tenant SaaS e-commerce platform where each seller operates their own isolated storefront with role-based data access. Built with Next.js 16 App Router and PayloadCMS as the headless backend.

• Integrated PayloadCMS with the multi-tenant plugin for declarative role-based access control; tenant users are scoped to their own data and super-admins have global access
• Built end-to-end type-safe data layer using tRPC 11 and Payload's auto-generated TypeScript types, with no manual type maintenance required
• Implemented secure httpOnly cookie authentication server-side; the JWT is never exposed to JavaScript, preventing XSS token theft
• Built infinite scroll product listings using TanStack Query’s useInfiniteQuery with server-side prefetching for fast initial load
• Hierarchical category system (parent/subcategory) powering nested navigation and multi-level product filtering
• Used React Hook Form + Zod for type-safe form validation shared across client and server
• Maintained feature branch Git workflow with pull requests reviewed by CodeRabbit AI for every merge
• Wrote a seed script to populate the database with tenants, products, and categories for development and demo purposes
• GitHub: github.com/IvanTran-2001/multitenant-ecommerce

Full-stack web application built with MongoDB, Express.js, React, and Node.js. Features secure user authentication and personalized note management with custom implementations of advanced security measures.

• Implemented JWT authentication with 7-day token expiration and bcrypt password hashing
• Designed MongoDB database schema using Mongoose for users and notes with relationship management
• Built RESTful API with Express.js supporting full CRUD operations
• Developed React frontend with protected routes and real-time user notifications
• Integrated Redis-based rate limiting to prevent abuse and ensure system stability
• Designed responsive UI using TailwindCSS and DaisyUI component library
• Deployed application on Render with MongoDB Atlas cloud database
• Live Demo: https://mern-thinkboard-qsrl.onrender.com

Engineered a feature-complete vending machine simulator in C++, applying low-level memory management, custom data structures, and established design patterns.

• Implemented custom linked list data structures from the ground up
• Applied the Command design pattern across 8+ operation classes for modular, extensible architecture
• Designed a change-dispensing algorithm supporting 8 coin denominations with optimal selection logic
• Implemented structured error handling with 4 custom exception types
• Persistent data storage via file I/O (stock.dat, coins.dat)
• Developed 16+ unit test cases covering edge cases and failure scenarios
• Zero memory leaks confirmed via Valgrind analysis

Developed a Node.js automation tool to scrape and validate article sort order on Hacker News, submitted as a technical assessment.

• Leveraged Playwright for cross-browser web scraping and DOM interaction
• Implemented data validation with structured error reporting and exit codes
• Clean CLI interface with configurable options
• Unit test suite written with Jest
• Modular architecture with reusable, single-responsibility components

Conducted a consulting engagement to analyse and re-architect the database layer of an existing warehouse management system.

• Performed requirements analysis and identified inefficiencies in existing workflows and schema design
• Re-architected the relational database schema for improved normalisation and maintainability
• Optimised SQL queries, reducing execution time and improving reliability
• Produced comprehensive technical documentation detailing schema changes and design rationale

200+ problems solved across different difficulty levels.

• 71% acceptance rate
• Top 17% runtime performance globally, demonstrating consistent focus on time-complexity optimisation
• Topics: Dynamic Programming, Backtracking, Hash Tables, Graphs, Trees, Arrays
• Languages: Python (121 problems), MySQL (3 problems)
• Easy (68), Medium (100), Hard (25)

• Gold Badge in React — 5-star, 600 points
• Frontend Developer (React) Role Certificate
• JavaScript Basic Certificate