Case study · Team project

CekPKWT

An AI agent that checks an Indonesian fixed-term work contract (PKWT) against current labour law before a fresh graduate signs it.

IBM LangflowIBM BobMCPRAGAstra DB
CekPKWT — Product shot — paste a contract, get a report

01 — Problem

Fresh graduates usually sign their first PKWT within days, with no way of telling which clauses break the rules: a probation period in a fixed-term contract, a wage under the regional minimum, a flat resignation fine. Unemployment for 15-24 year olds sits at 16.89%, over three times the national rate, so the bargaining power is all on the employer's side. A general chatbot does not help — it cites articles that were amended or struck down, with nothing the user can check.

02 — Approach

One Langflow flow serves three doors: IBM Bob over MCP, the public web app over the Langflow API, and CI. Personal data (NIK, NPWP, phone, bank number, name and address lines) is masked before any text reaches the LLM, and the contract is never written to the vector store. The agent extracts key clauses, retrieves articles and 2026 regional minimum wages from a 64-entry regulation corpus where every article carries its legal status and regime, and reports each finding with the article, its status, a severity and a polite question for HR. A daily bot reads JDIH and Constitutional Court rulings, rejects any text not found word for word in the official PDF, and opens a pull request that CI will not merge until a human reviews it.

03 — My role

Two-person team with Joel Chriscendo, built for an IBM Bob + Langflow challenge (Education & Future of Work).

Team project

Results

Live in production at a public URL, free and with no account, built from first commit to deployed web app and daily sync in one month. On the bundled sample contract it surfaces more than Rp12 million in wages and compensation the worker would otherwise lose. The sync bot has already opened real pull requests for new regulations (PP 49/2025, Permenaker 7/2026), and the agent cannot send email itself — the user reviews and sends every HR question from their own Gmail or WhatsApp.

64

Verified articles

1 month

First commit to prod

Rp12M+

At stake, sample contract

Screens
01 / 05
Tempel kontrak — personal data masked, with a preview before sending
Laporan — every finding cites its article and whether it is in force
Tanya aturan — follow-up answers only from the stored regulations
Draf HR — findings turned into a polite email the user sends themselves
The Langflow flow behind the contract check
Tempel kontrak — personal data masked, with a preview before sending

Demo recording

Pitch video — from pasted contract to a question for HR

Next projectDetektor Kata Kasar Indonesia