Research & Data
AI-powered economic research on climate, trade, and sustainable finance, producing working papers, policy briefs, and open datasets.
A lab where rigorous economics, social sciences, and data meet artificial intelligence, for Indonesia and the Global South.
AIEPA generates evidence-based insights that support sustainable and inclusive development, for Indonesia and the Global South.
We sit at the intersection of three disciplines: applied statistics for causal rigor, data science and machine learning for scale, and social-science inquiry (political economy, geopolitics, and institutional analysis) for contextual depth. Our work serves ministries, businesses, multilateral agencies, civil society, and the next generation of economists and social scientists training at UIII.
AI-powered economic research on climate, trade, and sustainable finance, producing working papers, policy briefs, and open datasets.
Capacity building for graduate students and professionals in AI, data science, and applied econometrics, through bootcamps and workshops.
Evidence-based policy analytics for governments, NGOs, and the private sector. We also provide an affordable data platform for researchers.
To become a leading interdisciplinary lab in Southeast Asia for AI-driven economics, social sciences, and policy research, bridging the gap between cutting-edge methods and real-world policy impact.
Grounded in evidence and strong academic tradition.
Research that serves society, not just academia.
Pioneering AI methods for economics and policy.
AI-powered climate risk assessment, emissions trading analysis, and green transition modeling for Indonesian policy.
Tariff impact analysis, trade policy simulation, export-control scenarios (IEEPA, supply-chain weaponization), and fiscal policy effectiveness studies.
Green sukuk analysis, ESG data analytics, and sustainable investment outlook reports for emerging markets.
Rigorous evaluation for development projects using difference-in-differences, RCTs, and quasi-experimental methods.
NLP on policy documents, parliamentary speeches, and media corpora; large language models for policy-relevant text, bridging AI methods and social-science questions.
Data science bootcamps, econometrics workshops, and hands-on AI training for graduate students and professionals.
Three tracks — quick snapshot analyses, policy-facing insights, and full working papers.
Early readouts and method demos we want to get into circulation quickly.
A citation map of every Indonesian law journal indexed in Scopus, set against the global Q1 universe. 346 journals, 17.3 million reference records. The picture: dense intra-network clustering, with high-ranked Indonesian journals citing each other at high frequency.
Indonesia's KBJI 2014 catalogs 2,136 occupations and, scattered across its entries, hundreds of "related occupations" notations that no team has ever extracted. AI did. The result is a 433-node, 656-edge network of occupational relatedness — and a striking finding: Green Potential occupations concentrate where the network is thinnest.
~95,000 games on the world's biggest PC marketplace, audited and analyzed. What wins, who plays where, and a 5-year forecast on platforms, prices, and play-modes — written for anyone curious about the gaming industry, not just insiders.
Indonesia is a coastline of 17,000 islands strung along the world's busiest maritime corridor, where roughly a third of all seaborne trade passes through the Malacca, Sunda, and Lombok straits each year. Trade Port is an interactive map of the network this geography produces — every international shipping connection between Indonesia's 70 active international ports and the 124 countries they serve.
China's Overseas Power Investment, an interactive atlas. Maps every debt and equity flow from Chinese banks and state-owned enterprises into overseas power plants between 1948 and 2030: 592 projects across 90 countries, totaling over USD 137 billion in committed finance.
One waterway carries a fifth of Indonesia's crude imports. We model what disruption does to the rupiah, the budget, and 2026 growth. Three numbers a finance ministry should be looking at this week.
An agentic-AI pipeline scores FOMC and Bank Indonesia communications on a common hawkish–dovish scale, revealing a lead-lag of roughly nine months (ρ ≈ 0.70) in how BI's tone tracks the Fed's.
Notes pitched to ministries, regulators, and multilateral partners.
Our first Policy Insights are in the pipeline and will land through 2026.
Full methodology, full results — work on its way to peer review.
More releases on the way. Policy Insights and additional Working Papers will be published on a rolling basis through 2026.
Subscribe for releases →AIEPA is working with Bank Mandiri's Office of the Chief Economist on a research engagement that rebuilds parts of an economist's daily workflow around AI. See the project →
Singapore's foreign minister open-sourced his personal AI assistant. We read it against ~25 comparator countries, lay out three risk registers, and propose a research agenda only AIEPA can run. Read analysis →
A tour of the Model AI Governance Framework, its GenAI addendum, and what Indonesia should think hard about before borrowing wholesale. With a four-jurisdiction comparison chart. Read analysis →
Anthropic's cybersecurity scare and the Glasswing coalition redrew the map of who gets to defend themselves. What that means for Indonesia and the Global South. Read analysis →
A model marketed as the ethical one still trains on the same scraped internet as the rest. What that means for anyone thinking carefully about AI use, with a look at the corpus composition. Read analysis →
Long-context models are quietly rewriting the empirical workflow. Plus OpenAI structured outputs, DeepSeek V3.2, and a reasoning-plus-executor pattern for cutting costs. Read analysis →
The AI, Economics, and Policy Analytics Lab formally opens its doors, bringing together economists, data scientists, and policy researchers under one interdisciplinary roof.
Short, practitioner-focused notes from the lab on working with frontier AI tools: what to pick, how to price it, and how to prompt it. Written for economists, social scientists, and policy teams getting hands-on with the stack.
An opinionated map of where Indonesian government data lives, what it takes to get at it, and where the legal and ethical lines sit for a research lab without institutional legal cover.
The Indonesian NLP toolbox in 2026 — IndoBERT, BGE-M3, NusaCrowd, Sahabat-AI — and the register problem that quietly breaks models on formal Indonesian text.
The 2026 agentic landscape has fragmented into terminal CLIs, IDE agents, VS Code extensions, and cloud platforms. The AIEPA stack, with cost discipline and a sovereign fallback.
Effective prompting techniques we've been using for data wrangling, code review, and literature synthesis in applied research. Worked examples included.
More partnerships coming soon.
Hands-on training for UIII and other Indonesian universities in AI, data science, and applied econometrics, supporting thesis research, publications, and early-career pathways.
Indonesian ministries, BAPPENAS, Bank Indonesia, OJK, and international bodies, delivering actionable analytics on trade, climate, and fiscal policy.
Corporations, financial institutions, multilateral agencies (World Bank, ADB, UNICEF), and NGOs seeking rigorous economic research and AI-powered insight.

Co-founder & Executive Director

Co-Founder & Co-Director
Lecturer, FEB UIII

Principal Scientist (Non-Resident)

Head of Research

Research Scientist

Research Scientist

Research Scientist

Research Scientist

Research Scientist

Senior Fellow

Senior Fellow

Senior Fellow

Senior Fellow

Senior Fellow

Senior Fellow

Senior Fellow