Applied AI
Models as one part of a product: a tutor that answers from the course material through RAG, and a ghost that learns a player's habits before it hunts them with A*.
Available for work and internships
Artificial Intelligence undergraduate and laboratory assistant at Binus University. I build whole systems, and now I'm pointing them at machine learning.

Most of what I have built so far is the part around a model: services talking over gRPC, a network that spans four sites, a cluster that heals itself, and a game whose ghost learns where you like to go. Now I study artificial intelligence, and I want the models I build to live inside systems that hold up.
I'm 19, in my third semester of Artificial Intelligence at Binus University, and I also work there as a laboratory assistant.
The job is teaching practicum classes, grading students' work, and writing the cases they solve. It also comes with self-development: each generation of assistants runs its own program, and mine, TPA 26-1, is where the five projects below come from.
Everything in the work section below belongs to one of these three.
Models as one part of a product: a tutor that answers from the course material through RAG, and a ghost that learns a player's habits before it hunts them with A*.
Services, data and the screens people use: Go microservices over gRPC, a Tauri desktop app with a Rust core, payments that never stop halfway, and real-time chat.
What everything else runs on: a four-site network joined by BGP, and a Kubernetes cluster provisioned by Terraform and Ansible and deployed through GitOps.
Five systems, each built from the architecture down to the last screen during TPA 26-1. Four of them I built alone.
An island's people, taught online by the best trainers it could find.
Online course platform built on microservices, with adaptive HLS video streaming, a RAG-powered AI tutor, payments, and livestreams.
View projectA cruise ship's whole crew, run from one desktop app.
Cruise-ship operations desktop app with role-based access, real-time chat and notifications, seat booking, and workload-based task assignment.
View project
A haunted port where some players hunt ghosts and the others are the ghosts.
Asymmetric multiplayer horror game of Exorcists versus Ghosts, with proximity voice chat, A* ghost AI that learns player habits, and encrypted saves.
View project
Four sites of an AI institute, joined into one secure network.
Multi-site enterprise network across four locations, using BGP, OSPF, EIGRP and RIP with HSRP, an ASA firewall DMZ, and an IPSec VPN.
View projectA real-time chat app, taken from a repo to a locked-down, self-healing cluster.
Production deployment of a React and NestJS chat app on a highly available Kubernetes cluster, with MikroTik networking, Proxmox, Terraform, Ansible and a GitOps CI/CD pipeline.
View projectI pick tools for the problem, not the trend. Everything here comes from the projects above, not from a reading list.
I am looking for machine learning or data analysis work. Email is the fastest way to reach me and I answer within a day.
michael15010407@gmail.comCVWeb version, with a PDFWhere
Jakarta, Indonesia · GMT+7
Studying
Artificial Intelligence, Binus University
Languages
Indonesian (native) · English (professional)