Skip to content
IDEAS
★ 63.5 / 100 • Rank #14 Originality: Incremental Adaptation Judge: Worth Prototyping AI Helpful Easy Future-Ready Workforce 30-Hour Build Plan

KontrataBantay: Multilingual OFW and domestic worker employment contract risk auditor

#Legal Tech #Tagalog AI #Livelihood #Computer Vision #SMS & Voice
Executive Elevator Pitch

"KontrataBantay protects Overseas Filipino Workers, domestic workers, and agency hires before they sign an employment contract. A worker photographs their English, Arabic, or bilingual contract; the app extracts every clause, compares it against the mandatory Department of Migrant Workers (DMW) Standard Employment Contract and Labor Code, flags illegal terms in red, and speaks a plain warning in Tagalog, Bisaya, or Ilocano."

Suggested Stack: Python FastAPI, Multimodal OCR (English/Arabic/Chinese/Tagalog), CLAUDETTE-style unfair clause classifier + DMW Standard Employment Contract RAG, Next.js PWA, Tagalog/Bisaya/Ilocano TTS.
Strict Judge Evaluation — 0–100 (Innovation 25 / Fit 25 / AI 20 / Feasibility 15 / Impact 15) Methodology & rankings →
Innovation & Creativity
56
Problem–Solution Fit
76
AI Application & Value
58
Feasibility & Potential
68
Impact & Scalability
58

Weighted total ★ 63.5/100 — ranked #14 of 75. Concept-stage assessment only: no prototype has been built or tested. Scores are comparative, not official PLDT marks.

Problem Statement

Over 2 million Overseas Filipino Workers (OFWs) and hundreds of thousands of local kasambahay and contractual agency workers leave their families to earn a living. Before departure, many are rushed into signing multi-page employment contracts written in dense legal English or bilingual Arabic/Chinese text, or are handed a second “substituted” contract at the airport or upon arrival abroad. These predatory contracts quietly insert illegal placement salary deductions, sub-minimum wages, 16-hour workdays without rest days, penalties for early termination due to abuse, or clauses permitting the employer to confiscate the worker’s passport—violating the Migrant Workers and Overseas Filipinos Act (RA 10022), the Department of Migrant Workers (DMW) Standard Employment Contract, and the Kasambahay Law (RA 10361).

Proposed Solution

KontrataBantay is a pre-signing contract auditor and agency verifier that fits in any worker’s phone:

  1. Multilingual Contract & Side-Agreement Scanner: The worker photographs all pages of their job offer, employment contract, or salary deduction form (even if written in English, Arabic, Mandarin, or Tagalog).
  2. Clause-by-Clause Statutory Audit (CLAUDETTE Pattern): KontrataBantay segments the document into individual obligations and compares each clause against the mandatory minimum provisions of the DMW Standard Employment Contract for that specific job category and destination country (or the Philippine Labor Code / RA 10361 for local hires). It flags four critical risks in red:
    • Illegal Clauses: Passport/ID confiscation, excessive placement fee deductions, waiver of right to seek embassy help, or ban on rest days.
    • Sub-Standard Wage / Hours: Wage below the DMW bilateral minimum (for example, $400 USD minimum for household service workers) or missing 8-hour continuous rest.
    • Contract Substitution Diff: If the worker scans their DMW-verified contract at PDOS and later scans a second contract handed to them at the airport or abroad, KontrataBantay runs a semantic diff across languages to expose every altered number or deleted protection.
    • Missing Mandatory Protections: Omission of free food/accommodation, repatriation at employer’s expense, or medical insurance.
  3. Spoken Dialect Explanation & DMW License Check: Reads out every red flag in conversational Tagalog, Cebuano, or Ilocano while cross-checking the recruitment agency’s name against the official DMW list of licensed vs. suspended/canceled agencies.

Target Market

  • Primary users: First-time and returning OFWs (household service workers, caregivers, seafarers, construction and hospitality workers), local kasambahays, and BPO/agency contractual workers.
  • Public-sector and NGO buyers: Department of Migrant Workers (DMW), Overseas Workers Welfare Administration (OWWA) Pre-Departure Orientation Seminar (PDOS) providers, LGU Public Employment Service Offices (PESO) migrant desks, and OFW advocacy groups (Center for Migrant Advocacy, Blas F. Ople Policy Center).
  • Community benefit: Stops contract substitution and debt bondage before a Filipino worker boards a plane or signs away their rights.

How AI Is Used

  • Cross-lingual document OCR and layout segmentation: Reads multi-column bilingual contracts (such as English-Arabic unified domestic worker contracts) and aligns parallel clauses to check if the foreign-language column matches the English column.
  • Unfair and illegal clause detection (CLAUDETTE architecture): Classifies each clause against statutory prohibitions and unfair-contract taxonomies, citing the exact DMW Standard Employment Contract section violated.
  • Cross-document semantic diffing: Compares a worker’s original DMW-approved contract against a newly presented side contract to catch disguised wage cuts or altered job duties.

Community & Ecosystem Integration

  • Plugs into OWWA Pre-Departure Orientation Seminars (PDOS): Gives every departing OFW a 30-second tool to check any paper they are asked to sign at the agency or airport.
  • Live DMW agency license lookup: Checks whether the recruiter is valid, suspended, or on the DMW illegal recruitment blacklist.
  • One-tap DMW / Embassy emergency packet: If a worker abroad is handed a substituted contract, they can send the encrypted comparison report directly to the Migrant Workers Office (MWO / formerly POLO) and DMW helpline.

Demo Scenario (3 minutes for judges)

  1. Place a realistic 3-page bilingual (English/Arabic) Household Service Worker employment contract on the judge table containing three hidden traps: a $280/month salary (below the $400 DMW minimum), a clause stating the employer holds the passport “for safekeeping,” and a 6-month salary deduction for placement fees.
  2. Snap the pages with a phone. Within 5 seconds, KontrataBantay highlights all three illegal clauses in red on screen.
  3. Play the spoken Ilocano and Tagalog voice warning: “Babala: Nakalagay sa Section 6 na kukunin ng amo ang passport mo. Bawal ito sa ilalim ng DMW Standard Employment Contract. Ang nakasulat ding sahod ay $280 lang, mas mababa sa $400 minimum.”
  4. Also show the Bilingual Mismatch Detector catching a discrepancy where the English column says “1 day off per week” while the Arabic column says “1 day off per month.”

30-Hour Build Plan

Hours Task Role
0 to 4 Compile DMW Standard Employment Contracts (HSW, Skills, Seafarer), RA 10022, RA 10361, and DMW agency status list All
4 to 11 Build multilingual OCR + CLAUDETTE-style unfair clause classifier grounded in DMW mandatory rules AI / Legal NLP
4 to 12 Build bilingual column consistency checker (English vs. Arabic/Mandarin clause equivalence) and contract diff engine Backend / AI
11 to 18 Build DMW recruitment agency license lookup and MWO complaint PDF packet generator Backend
12 to 22 Build mobile camera scanner UI with red/amber/green clause highlights and side-by-side diff view Frontend
18 to 25 Wire Tagalog, Cebuano, and Ilocano TTS voice explanations for each flagged clause AI + Frontend
25 to 30 Test on 15 sample OFW and Kasambahay contracts with planted illegal clauses; rehearse 3-minute demo All

Suggested stack: Python FastAPI, Gemini 1.5 Pro / Flash Multimodal (for English/Arabic/Chinese OCR & statutory RAG), PostgreSQL, Next.js PWA, Edge/Google TTS. Public data / APIs: Department of Migrant Workers (DMW / POEA) Standard Employment Contracts and Governing Board Resolutions, DMW public database of licensed and blacklisted recruitment agencies, Republic Act No. 10022 and Republic Act No. 10361.

Research & Literature References

  1. Marco Lippi, Przemysław Pałka, Giuseppe Contissa, Francesca Lagioia, Hans-W. Micklitz, Giovanni Sartor, and Paolo Torroni, “CLAUDETTE: an automated detector of potentially unfair clauses in online terms of service,” Artificial Intelligence and Law, Vol. 27, Issue 2, pp. 117-139, 2019. DOI: 10.1007/s10506-019-09243-2
    • How KontrataBantay uses this study: Adapts the CLAUDETTE sentence-level legal classification and unfairness taxonomy pipeline to automatically spot unlawful, one-sided, or statutorily void clauses in worker employment contracts.

Risks & Privacy Notes

  • Worker identity and passport protection (RA 10173): Passport numbers, worker names, and residential addresses on scanned contracts are automatically redacted on-device before clause analysis, and images are never stored on public servers without explicit user opt-in for an official DMW complaint.
  • Clear legal disclaimer: KontrataBantay provides automated compliance comparisons against published DMW Standard Employment Contract templates and refers workers directly to DMW Legal Assistance / PAO lawyers for formal action.

AI Use Disclosure (draft)

KontrataBantay uses multimodal OCR and a retrieval-augmented LLM following the CLAUDETTE unfair-clause detection framework to compare contract clauses against official DMW Standard Employment Contracts and cross-check bilingual text columns. AI coding assistants helped build the document diff UI and agency lookup API. All statutory rules were verified against official DMW regulations.

Rubric Scoring

Criterion Weight Score (1-5) Why
Innovation & creativity 25% 3 Technically a multimodal OCR and RAG prompt wrapper over DMW standard employment contracts
Problem–solution fit 25% 4 Targets illegal contract substitution and excessive placement fees faced by departing OFWs
AI application & value 20% 3 Bilingual Arabic/English OCR and clause comparison use VLMs, though wage floors are simple lookup rules
Feasibility 15% 4 Very buildable in 30 hours using sample contracts and DMW country rules
Impact & scalability 15% 3 Helpful pre-departure check for OFWs, though legal enforcement still depends on DMW officers
Weighted total 3.40

More in Future-Ready Workforce

View all in pillar →
© 2026 PH AI Ideas
75 concepts / 6 pillars