← Back to work

AI · Case study · Private

Policy Lab

Laboratorium Kebijakan Indonesia - an AI-assisted lens that turns any public policy question into a structured, debatable, sourced analysis.

LovableLovable AI GatewayIndonesian-first prompts
internal
Policy Lab preview

Context

The setup before I touched it.

Indonesian public discourse around policy is loud but shallow. A new tax rule, a fuel subsidy change, a minimum wage debate - it shows up as a headline, a viral post, a 30-second TikTok. The boring middle (who actually pays, what the second-order effects are, what comparable countries did) gets skipped. Citizens form opinions on vibes; researchers waste hours framing before they get to thinking.

How it came together

Step by step - expand any phase for the highlights.

  • Pulled recurring policy debate patterns
  • Identified the missing 'boring middle'
  • Catalogued local context gaps in English tools
Listen to the discourse preview

Problem → Approach → Result

The short version, for the broad audience.

Problem

  • 01Policy debate is dominated by opinion and identity, not structured trade-off analysis.
  • 02Citizens have no accessible tool to stress-test the claims being made about policies that directly affect them.
  • 03Researchers and journalists spend the first half of every project on framing - mapping stakeholders, listing trade-offs, finding precedent - before the actual analysis begins.
  • 04Indonesian-language policy tools barely exist; everything assumes English and Western context.

Approach

  • 01Built a focused workspace where any policy question becomes a structured brief - stakeholders, costs, beneficiaries, risks.
  • 02Used AI to surface trade-offs, comparable cases from other countries, and steelmanned counter-arguments the user hadn't considered.
  • 03Kept the human firmly in the loop: every output is editable, sourced, and challengeable - the tool drafts, the human decides.
  • 04Indonesian-first by design: tone, examples, references, and historical precedent all rooted in local context, not translated from English.

Result

  • 01What used to take a researcher half a day (mapping stakeholders, finding comparable cases) now takes 10 minutes.
  • 02Conversations move from 'I feel this is unfair' to 'here's who pays, here's who benefits, here's the trade-off' - which is a much more productive argument.
  • 03A repeatable structure for thinking about new policies as they emerge, instead of reinventing the analytical frame every time.
  • 04Used as a personal sandbox by policy-curious citizens, not just professionals - which was the point.

Role

Built and maintained as a side project - product, design, and prompt engineering.

Status

Live - internal

Next

Browse more work →

React

Tap if this resonated.

What do you think?

Leave a comment or ping me — any feedback, thoughts, or collab ideas, I'll really appreciate it. Building alone is no fun.