I’m a Full-Stack Developer based in Lubuk Linggau, Indonesia, with a Bachelor’s degree in Informatics Engineering from Universitas Bina Darma (GPA 3.72). My focus is backend development with Node.js, Express.js, and NestJS, backed by solid relational database design in MySQL and PostgreSQL, plus working knowledge of React, Next.js, and TypeScript on the frontend.
Career so far. From December 2025 to June 2026, I worked as an IT Intern at Balai Pemasyarakatan Kelas II OKU Induk, where I was one of two developers building BAPAS OKU CONNECT, an internal web application. I took part in the full cycle: requirements analysis, design, implementation, testing, and debugging. I designed the relational database (tables, keys, relationships), built backend CRUD modules with Node.js, Express.js, and MySQL, and implemented JWT authentication with 15-minute access tokens, 7-day refresh tokens, automatic refresh, and HttpOnly cookies. I also implemented and debugged role-based access control, testing permissions and forbidden endpoints in Postman across Admin and Operator scenarios. Beyond my own modules, I reviewed and tested other developers’ features, reported bugs, and helped fix or improve them. I presented the application and its technical aspects to staff and leadership; it is currently at the development/staging stage, pending deployment approval. Earlier, in 2023, I was a Programmer Intern at AddOn Finance, where I studied database structure and server architecture in a manufacturing information system.
Notable projects. AgroLedger is a sales and finance management app for agriculture that I built solo in about a month. It automates settlement calculations for rubber and palm oil sales, including profit sharing, commissions, and deductions for credit and cash advances. For a transaction involving about 10 workers, it reduced calculation time from 1–2 hours of manual work to roughly 5–10 minutes and lowered the risk of human error. It’s built with NestJS, PostgreSQL, and Prisma, and includes an end-to-end OCR feature that reads transaction document photos, fills in the form, and asks for validation before saving. I kept the business logic in the service layer and cross-checked results against real transaction data before using it. My undergraduate thesis compared YOLOv8s and Faster R-CNN for detecting oil palm fruit maturity across five classes, using a 4,578-image dataset.
Interests. I enjoy building software that solves practical, real-world problems, especially systems with clear business rules, secure authentication and authorization, and clean backend architecture. I’m also interested in applying computer vision and OCR to automate manual work. I’m looking for backend or full-stack roles where I can keep growing, contribute to production systems, and work with a team that values quality and learning.
GitHub: github.com/Sakbanullah
