Problem Statement
TUPAD pays for 10-15 days of cleanup then ends; skilled habagat survivors return to zero. Meanwhile contractors rebuilding roads, schools, and GenSan fish ports can’t find vetted helpers. TESDA courses exist but workers don’t know which 3-day module unlocks the next paid gig.
Proposed Solution
Log TUPAD work + skills (photo of finished gutter, catch handled, tools used): HanapRebuild builds a verified micro-passport, matches to nearby rebuild gigs by distance/skills/pay, and links the exact TESDA micro-course that raises pay. SMS alerts work with no data; cooperatives post gigs with fixed pay to kill palakasan.
Target Market
- Primary users: TUPAD beneficiaries, displaced fishers, out-of-school youth in Bulacan, Pampanga, Cebu, GenSan.
- B2G: PESO, DOLE, TESDA field teams.
How AI Is Used
- Skills extractor: LLM turns work photos + Tagalog notes into verified skills.
- Matcher: Ranks gigs by fit, distance, and pay fairness.
PLDT / Smart / ePLDT / Rokid Angle
- Smart: SMS job matches for no-load workers.
- ePLDT: Hosts passport vault for PESO verification.
- Rokid: Trainers see hands-free guides on-site (stretch).
Demo Scenario (3 minutes for judges)
- Mock worker logs cleanup + tuna handling; passport appears.
- Three rebuild gigs rank with pay and distance.
- TESDA link shows “+₱150/day after 3-day course.”
- SMS job alert arrives on basic phone.
- Close: “Hindi isang ayuda lang.”
30-Hour Build Plan
| Hours | Task | Role |
|---|---|---|
| 0–3 | Mock gigs + TESDA modules | All |
| 3–12 | Extractor + matcher | AI lead |
| 3–10 | Passport + SMS | Backend |
| 10–20 | Worker UI | Frontend |
| 20–26 | Fairness testing | All |
| 26–30 | Pitch, disclosure | PM / Pitch |
Suggested stack: Next.js, FastAPI, LLM skills extractor, matching ranker, SMS alerts. Public data / APIs: DOLE TUPAD guides, TESDA course lists (public).
Risks & Privacy Notes
- Exploitation: Fixed-pay posting + report button; no placement fees. Demo workers fictional; no real IDs into AI tools.
- Privacy: Work data minimized, deletable, consent-first.
AI Use Disclosure (draft)
Uses LLM extraction plus matching. Coding assistants helped with UI. All demo matches human-checked.
Rubric Scoring
| Criterion | Weight | Score (1–5) | Why |
|---|---|---|---|
| Innovation & creativity | 25% | 4 | TUPAD-to-career passport is fresh |
| Problem–solution fit | 25% | 5 | Post-disaster jobs cliff is real |
| AI application & value | 20% | 4 | Extraction + matching need AI |
| Feasibility | 15% | 4 | Very demoable |
| Impact & scalability | 15% | 4 | Nationwide PESO use |
| Weighted total | 4.30 |