Aryo D. Prasetia — Systems Architect

How I Build

From Idea to System

XR_01

view_in_ar

XR Systems Engineering

▾

Build Steps

01Define interaction model & spatial constraints before touching engine

02Prototype core loop in Unity/Unreal — isolate performance bottlenecks early

03Build modular scene architecture — swappable assets, scalable logic

04Optimize render pipeline, draw calls, LOD strategy per target platform

05Integration testing across headsets — latency, comfort, fidelity balance

What You Get

→Production-ready XR app with documented architecture

→Performance budget defined and respected

→Team that understands the codebase and can extend it

Unity Unreal OpenXR WebXR

RT_02

precision_manufacturing

Real-time Simulation

▾

Build Steps

01Define simulation fidelity requirements — what must be real, what can be approximate

02Set up physics layer — PhysX, custom solvers, or hybrid approach

03Build lighting model — baked vs real-time tradeoffs per use case

04Instrument telemetry: frame time, simulation tick, render budget

05Stress test with edge conditions — not just happy path

What You Get

→Stable real-time environment with defined performance floor

→Physics accuracy matched to your actual requirements — no over-engineering

→Robotics/industrial simulation ready to connect to external systems

PhysX ROS2 Shader Graph HDRP

BE_03

database

Scalable Backend Systems

▾

Build Steps

01Map data flow first — where does it originate, how does it move, what reads it

02Define service boundaries — avoid premature microservices, avoid monolith traps

03Design API contracts before implementation — schema-first development

04Build observability in from day one — logs, metrics, traces

05Load test before production — not after something breaks

What You Get

→System that handles growth without emergency rewrites

→Infrastructure that your team can maintain and understand

→Documented architecture decisions — not just code

Node.js Go PostgreSQL AWS/GCP

WB_04

web

Web Platforms & Clean Architecture

▾

Build Steps

01Establish design system before writing component code

02Define state model — predictable data flow reduces entire classes of bugs

03Architect for change — every feature will be modified; build for that

04Integrate cloud services with abstraction layers — avoid lock-in

What You Get

→Platform that doesn't slow down as it grows

→UI that's consistent without being rigid

→Clean separation of concerns that new devs can navigate

React Next.js TypeScript Vercel

MN_05

school

Technical Mentorship

▾

How I Mentor

01Diagnose where the team is — skills gap, process gap, or confidence gap

02Pair on real work — not toy examples. Actual production problems

03Teach the why, not just the how — patterns that transfer across projects

04Build feedback loops — code review culture, architecture reviews

05Exit when the team can do it without me — that's the goal

What You Get

→Engineers who think in systems, not just features

→Team culture that multiplies knowledge instead of hoarding it

→Structured game dev / coding curriculum if needed

1-on-1 Team Workshops Code Review Game Dev

        Not every problem needs a six-week architecture sprint.

        Sometimes the best move is: build it in a weekend, learn what's real, then decide.

        I run rapid prototypes to validate ideas fast — especially with new tech I want to understand deeply.

⚡ RAPID · AI INTEGRATION

AI-Powered Features

Spinning up LLM integrations, vision models, or generative pipelines to test feasibility — before committing to a full product architecture. I explore AI not just as a tool user but as someone who wants to understand the stack underneath.

OpenAI API LangChain Hugging Face Local LLMs

⚡ RAPID · XR PROOF OF CONCEPT

XR Concept Demos

Got an XR idea but not sure it's viable? I can put a working spatial prototype in front of users in days — not months. Fast enough to validate before a single hire or budget line is committed.

Unity Meta SDK WebXR

⚡ RAPID · NEW TECH EXPLORATION

Tech Experiments

When a new framework, engine update, or platform drops — I build something real with it. Not tutorials. Actual problems. This is how I stay current and how I know what advice to give teams considering adoption.

Emerging APIs Beta SDKs Edge Cases

⚡ RAPID · BACKEND SPIKE

Backend Spikes

Need to know if a real-time multiplayer feature, a data pipeline, or a cloud integration is actually feasible at your scale? I spike it in isolation — no assumptions, real numbers — so decisions are made on evidence.

WebSockets Redis Load Testing

⚡ RAPID · GAME PROTOTYPE

Game Mechanic Prototypes

A mechanic isn't real until it's playable. I prototype game ideas fast — the loop, the feel, the edge cases — and iterate until it's worth building properly. Jam-style speed, production-grade thinking.

Unity Godot Game Jam

⚡ RAPID · AI + XR

AI Inside XR

Combining LLMs, computer vision, or generative models with spatial computing. NPC behavior, spatial understanding, real-time inference inside a headset. This is the frontier I enjoy the most.

Vision Models Spatial AI NPC AI Edge Inference

I Build the Structure
Behind the Feature.

From Idea to System

System Capabilities

Let's build something
worth shipping.

I Build the StructureBehind the Feature.

From Idea to System

System Capabilities

Let's build somethingworth shipping.

I Build the Structure
Behind the Feature.

Let's build something
worth shipping.