Analisis Intervensi & Anggaran
Structured Problem Decomposition & Budget Planning System
- Role
- Front-End & Systems Integration Engineer
- Period
- 2024 – 2025
- Organisation
- Kementerian PPN/Bappenas – SKALA Program
Overview
A structured policy planning application connecting problem identification, root-cause analysis, intervention options, and budget estimation into a unified decision-support workflow.
Context
Regional poverty and basic service deficits require clear causal chains: identifying the core issue, finding root causes, selecting proven intervention packages, and allocating realistic local budgets.
Problem
Regional budgets frequently misalign with the real root causes of poverty due to fragmented planning steps and lack of standardized intervention models.
Approach
Developed a guided multi-step planning pipeline that enforces causality (Masalah → Penyebab → Intervensi → Anggaran), enhanced with AI-assisted problem-tree structuring.
Flow
- Masalah
Core Problem
Select and define the focal regional deficit
- Penyebab
Root Causes
Decompose systemic causes using historical and statistical evidence
- Intervensi
Intervention Options
Select evidence-based policy packages tailored to local context
- Anggaran
Budget Allocation
Simulate regional fiscal costings and calculate funding requirements
Technical challenges
- Translating complex recursive problem-tree data structures into a responsive, fluid interactive UI.
- Integrating asynchronous AI/RAG response pipelines for real-time problem-tree recommendations without blocking the main workflow.
- Ensuring regional budget calculations dynamically update across multi-tiered intervention items.
Engineering decisions
- Decoupled the tree-state graph management from presentation nodes to allow instant re-parenting and editing.
- Built resilient optimistic UI updates with debounced API synchronizations to withstand intermittent government network environments.
- Collaborated on API contracts with the backend/AI team to stream structured JSON nodes seamlessly to the front-end parser.
What shipped
- Strict problem-to-budget causal pipeline workflow
- AI-assisted problem tree decomposition powered by RAG integrations
- Real-time regional budget estimation simulator
Stack
React.js · Next.js · TypeScript · Tailwind CSS · Python / FastAPI · AI / RAG Integration