Fullstack Engineer — Intelligent & Data Systems
Building scalable platforms powered by AI, spatial intelligence, and data-driven architecture.
Production platforms deployed for government agencies, industry operators, and civic infrastructure across Indonesia. Each entry documents the problem, approach, architecture, and real-world impact.
↓ Expand each card to see the full case study.
Kabupaten Klaten had no unified operational intelligence layer. Agricultural monitoring, satellite imagery (NDVI), disaster risk data, plant disease detection, and government service APIs all existed in separate silos and making real-time decision-making for local government nearly impossible.
Indonesia's national sugar self-sufficiency (Swasembada Gula) program required a coordinated digital layer across PT Sinergi Gula Nusantara's operations. Plantation data, mill performance, sugarcane farmer management, and reporting to headquarters were fragmented across dozens of PTPN entities without a unified system.
Sugarcane plantation operators in Indonesia were making field-level decisions with outdated, manual data. There was no way to query spatial health indices, yield predictions, or operational status at the plantation polygon level in real time.
Manual verification of palm oil plantation boundaries was expensive, slow, and error-prone at national scale and creating significant compliance gaps in Indonesia's palm oil regulatory system managed by Ditjenbun (Directorate General of Plantations).
Knowledge workers at digii.co.id needed an AI-assisted layer for data management and analysis a way to query, summarize, and extract insights from operational data using natural language rather than manual SQL or spreadsheet analysis.
Indonesia's Directorate General of Plantations (Ditjenbun) needed centralized, real-time visibility over palm oil commodity data across all 34 provinces. Existing reporting was fragmented, delayed, and incompatible across provincial systems.
Principles I apply when designing and building systems not a methodology, but a consistent way of reasoning about problems.
Every feature request is first evaluated as a systems problem. What are the dependencies? What are the failure modes? What does this look like at 10x scale? I rarely build for today's requirements alone.
I trace data from source to display. Collection, normalization, storage, query, presentation and where each step introduces latency, error, or inconsistency. Most bugs live in the gaps between stages.
Every architectural decision trades something: consistency vs. availability, simplicity vs. flexibility, build vs. buy. I surface these trade-offs explicitly and document the reasoning not just the decision.
Code that works on localhost is not done. Deployment story, error handling, observability, role-based access, and maintainability are part of the spec not afterthoughts.
My background spans full-stack development, AI inference, and geospatial data but the consistent thread is systems thinking applied to real operational problems. I'm drawn to problems where architectural decisions have outsized impact on outcomes not just correctness, but performance, maintainability, and operator trust.
Three intersecting domains where I have built and deployed production systems — not surface-level familiarity, but genuine technical depth backed by real-world deployments.
Production AI — not demos. From computer vision pipelines processing satellite imagery for government use, to self-hosted LLM inference (Ollama) embedded into operational dashboards, to RAG-based agent platforms for internal productivity. The focus is always on deployability, latency, and accuracy under real-world conditions.
Capabilities
Proof of work
• KLIP: self-hosted Ollama LLM for civic data queries
• Palm Oil Detection: CV pipeline for PalmCo (govt)
• Agent AI: RAG-based productivity platform at agentai.digii.co.id
• NDVI satellite analysis with Python/Flask microservices
• Palm Counting: CV system for plantation yield estimation and forecasting
Spatial data as a first-class engineering concern. Building GIS platforms for government agencies and industry operators processing satellite imagery, serving NDVI tile layers, running PostGIS spatial queries, and rendering complex geographic data at operational scale. Applied across sugarcane, palm oil, and civic smart city contexts in Indonesia.
Capabilities
Proof of work
• Montera: plantation GIS for sugarcane industry operators
• KLIP: real-time NDVI tile rendering with Mapbox GL JS
• Etera: multi-tenant plantation management GIS for PTPN Group
• E-STDB: national palm oil monitoring dashboard with geospatial data layers
Platforms where data quality and system reliability directly affect public or operational outcomes. Government-deployed systems for agricultural monitoring (Ditjenbun, PTPN Group), disaster response (KLIP smart city), health screening, and national commodity tracking plus enterprise dashboards for KPI monitoring, HR analytics, and asset certification. Built for durability and operator trust, not for demos.
Capabilities
Proof of work
• E-STDB Dashboard: national palm oil monitoring (Ditjenbun)
• ETERA: PTPN Group national sugar industry platform
• Simkeswa: maternal health screening system
• KPI SGN, People Analytics, AMS Certification
• People analytics system for internal HR insights
• Montera: plantation GIS with real-time operational data for field managers
Fullstack engineer with 5+ years building data-driven platforms for government, agriculture, and enterprise in Indonesia. Currently at PT LPP Agro Nusantara (PTPN Group) as IT Fullstack Developer contributing to national sugar industry digital infrastructure. I specialize at the intersection of AI inference, geo-spatial data, and operational software: systems where architectural decisions have direct real-world impact.
My work spans full-stack development, AI inference pipelines, and geospatial data deployed for government agencies (Ditjenbun, PTPN Group, local government) and industry operators. I care about systems that work reliably under real conditions, not just systems that pass the demo.
Currently open to remote engineering roles and technical collaborations — particularly in AI products, data infrastructure, and civic technology.
IT Fullstack Developer
PT LPP Agro Nusantara (PTPN Group)
Building ETERA platform for national sugar industry digital infrastructure
Fullstack Developer
Mariban
Maintained and optimized retail web systems; performance tuning
Fullstack Developer
Gotara Indonesia
Engineered full-stack farm management systems for modern livestock operations
IT Support Specialist
PT Global Jet Cargo
Managed enterprise web systems for logistics operations
Open to remote engineering roles, contract engagements, and technical collaborations. If you're working on something in AI, data infrastructure, or civic technology. Let's talk.
hapisadi12@gmail.com