Decidix: A Bilingual, Zero-Cost Grounded AI and Decision Support Platform for Regional Development Analytics
Public-sector development planning and regional resource allocation frequently suffer from fragmented data silos, delayed reporting cycles, opaque predictive models, and prohibitive artificial intelligence operational costs. This paper presents Decidix, a production-grade, bilingual (Arabic RTL / English LTR) Decision Support System (DSS) engineered across web (React/TypeScript) and mobile (Flutter) clients backed by a cloud-native Node.js/PostgreSQL architecture. Decidix unifies four core development sectors—Education, Healthcare, Economy, and Infrastructure—into a single analytical platform featuring a Unified National Dataset Registry with atomic versioned ingestion, a Zero-Cost Grounded RAG Assistant utilizing deterministic 1536-dimensional local embeddings with pgvector cosine similarity search, and a Concave Diminishing-Returns What-If Policy Simulator. Empirical verification demonstrates high system reliability across 251 Flutter automated tests, 24 web frontend unit tests, 72 multi-viewport browser QA assertions, and over 240 backend unit and live PostgreSQL integration tests, operating with $0.00 in external AI API expenditure.
Authors
- Mohanned Mohamedali (ORCID: https://orcid.org/0000-0003-2491-5697)
Institutions
- University of Calgary (CA)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-21
- DOI
- https://doi.org/10.5281/zenodo.24416458
- Primary Topic
- Decision Support System Applications
- Type
- article
- Field-Weighted Citation Impact
- 0.00