Workforce · Vision AI
SkillPass
An AI-verified skills passport for the world's migrant workers.
Workers perform a practical task on video at a partner training centre. AI scores the recording against a trade rubric, a human assessor signs off, and the worker gets a verifiable passport that employers abroad can trust — translated into their language. Better-matched workers earn more and are cheated less.
Launch market: Bangladesh · 5 cited sources · Facts reviewed 14 October 2026
- Cash to MVP (est.)Our estimate of the cash needed to reach a working MVP: one-time setup plus monthly running costs for the build period. It excludes founder salaries.
- $3,370
- MVP timelineMinimum viable product: the smallest version real users will use and pay for, built to learn fast.
- 8 weeks
- Margin after AI (est.)Of every 100 you charge, how much is left after paying for the AI models and infrastructure behind it. Higher means a healthier business.
- ~60%
- Editorial scorePlankito's own judgement: the average of six 1–5 ratings (AI leverage, moat, competition gap, capital efficiency, timing, scalability). It is an opinion to compare ideas, not a forecast.
- 4.2 / 5
At a glance
- Who uses it
- Workers going abroad, and the training centres that prepare them.
- Who pays
- Employers abroad, per verified hire; training centres pay for the software.
- Who should build it
- A full-stack engineer plus someone with recruiting or training-centre networks.
- The hardest part
- Making employers trust an AI-verified score enough to hire on it.
01
Opportunity
Workers going abroad are hired on paper claims, so employers cannot tell a skilled electrician from a beginner. Skilled workers are paid as unskilled, employers face mismatches on arrival, and workers carry the risk and the debt. Destination countries are now testing skills — and many trained applicants fail.
The evidence
167.7M
Globalinternational migrants work in their destination countries' labour forces — 4.7% of the global workforce.
1.13M
Launch marketBangladeshis went abroad for work in 2025, up 12% on 2024; 90% went to just five countries.
78%
Launch marketof Bangladeshi migrant workers in 2025 were classed as less-skilled or semi-skilled; only 2.9% professional.
~20%
Launch marketof trained Bangladeshi applicants failed Saudi Arabia's skill verification test since mid-2023.
Why now
- Vision models can now follow a recorded practical task step by step and score it against a rubric.
- Translation makes one profile readable to employers in Arabic, English and other languages instantly.
- Saudi Arabia now requires skill verification tests for many trades, and one in five trained Bangladeshi applicants fails them.
- Remittances hit a record $35.6 billion in Bangladesh's FY2025–26, so better-paid migrant jobs move the whole economy.
02
The AI case
What AI makes possible here, why the big labs won't absorb it, and whether the margins hold.
AI leverage — what is newly possible
- Scores a recorded practical task — wiring, welding, tiling — step by step against the trade rubric.
- Spots safety violations in the video that a rushed human assessor might miss.
- Generates an employer-ready profile in Arabic and English from the assessment and the worker's own words.
- Recommends targeted re-training before a destination-country test, focused on the steps the worker failed.
Lab-proof moatsA reason competitors, including big AI companies, can't easily copy you: data, distribution, trust or local knowledge.
Labelled assessment video
Thousands of scored practical tasks per trade, labelled by certified assessors. No public dataset like it exists.
Two-sided network
Training centres bring workers; employers bring jobs. Each side makes the passport more valuable to the other.
Verification trust
Signed, revocable credentials with a fraud-control record. Trust is earned slowly and is hard to copy.
Inference economicsThe cost of running AI models each time a customer uses the product. It decides whether an AI business makes money. (estimates)
- Unit
- Verified assessment
- AI cost per unit
- ~$0.60 AI + ~$4 human assessor
- Price per unit
- ~$12–15 (paid by training centre or employer)
- Gross margin
- ~60%
How the margin is protected
- →AI pre-scoring lets one assessor sign off far more tests per day.
- →Charge employers per verified hire, not per profile view.
- →Re-training recommendations become a paid product for training centres.
03
Market & expansion
Launch in Bangladesh. Then expand.
- Three customers: training centres want placement rates, employers abroad want verified skill, workers want better jobs.
- Start with two trades in high demand, such as electrical and welding, and one destination country.
- Workers must never pay the platform; revenue comes from employers and institutions.
- Validate: will employers pay per verified hire, and which certifications they already trust.
Editorial scorecardPlankito's own judgement: the average of six 1–5 ratings (AI leverage, moat, competition gap, capital efficiency, timing, scalability). It is an opinion to compare ideas, not a forecast.
- AI leverage
- 4/5
- Moat durability
- 4/5
- Competition gap
- 4/5
- Capital efficiency
- 4/5
- Timing
- 5/5
- Scalability
- 4/5
Where it travels next
Nepal, India, Pakistan, Sri Lanka
Major labour-sending countries hiring into the same Gulf employers.
Philippines & Indonesia
Large, well-organised migration systems that value skill verification.
East & West Africa
Fast-growing outbound labour to the Gulf and Europe, with few trusted skill signals.
04
Competition & edge
Who already plays here, what they leave open, and what could go wrong.
Recruiting agencies
Licensed agencies that place workers.
GapIncentivised by volume, not verified skill; little transparency for workers.
Government skills certificates
National qualifications and test centres.
GapPaper-based and slow for foreign employers to verify.
Global job boards
General hiring platforms.
GapNot built for blue-collar migration or practical skill proof.
Where the edge comes from
- Video proof of practical skill, which no paper certificate offers.
- Partnership with existing training centres instead of building schools.
- Worker-first trust: free for workers, which is rare and marketable.
Risks and how to handle them
AI scoring is biased or wrong.
AI pre-scores, a certified human signs off; audit AI–human agreement per trade every month.
Fraud in assessments.
Identity check on camera, assessor sign-off, signed credentials and random re-tests.
Recruiting agencies see it as a threat.
Offer agencies a premium verified-candidate pool rather than competing for placements.
05
Business model
Employer fee
Per verified hire or annual subscription to the talent pool.
Training centre SaaS
AI assessment, re-training plans and placement reporting.
Programme contracts
Outcome tracking for governments and development partners.
06
Stack & data
Modern, proven tools a small team can run without a DevOps hire.
Web app
Next.js 16 (App Router) + React 19
Employer portal, centre dashboard and public verification pages from one codebase.
Styling
Tailwind CSS v4 + shadcn/ui
Professional, trustworthy UI quickly.
Video
Mux or Cloudflare Stream
Upload from low-end phones; adaptive playback for employers.
Assessment AI
Multimodal frontier model scoring sampled frames against a structured rubric
Step-level scoring and safety checks with explanations an assessor can verify.
Profiles
Claude API for translation and employer-ready summaries
Accurate Arabic and English profiles from Bangla input.
Database
PostgreSQL (Neon) + Drizzle ORM
Relational records with strict audit trails.
Verification
Signed credentials with QR codes and revocation
Anyone can verify a passport in seconds.
Hosting & ops
Vercel + Sentry + PostHog
Simple, reliable operations.
Data model6 core tables, for the engineer starting the first migration.
workers
Worker identity and profile.
- id
- phone
- name
- nid_hash
- district
- languages[]
trades
Skill categories and rubrics.
- id
- name
- rubric (json)
- level
assessments
A recorded practical test.
- id
- worker_id
- trade_id
- centre_id
- video_id
- ai_score
- assessor_score
- passed
rubric_scores
Step-level scores.
- assessment_id
- step
- ai_score
- assessor_score
- notes
passports
Public verified profile.
- id
- worker_id
- public_slug
- signature
- issued_at
- revoked_at
placements
Hiring outcomes.
- id
- worker_id
- employer_id
- salary
- start_date
- retained_90d
07
Costing
Directional monthly estimates in USD. Your numbers will differ — the structure is what matters.
| Item | MVP / monthWhat the product costs to run each month (hosting, AI models, tools) at MVP stage. | Scale / month |
|---|---|---|
| Hosting (Vercel Pro) | $20 | $150 |
| Database | $25 | $200 |
| Video hosting & streaming | $30 | $700 |
| Assessment & translation AI | $60 | $900 |
| Human assessor feesPaid per assessment | $300 | $2,500 |
| Monitoring & analytics | $0 | $100 |
| Total | $435 | $4,550 |
One-time setup
08
MVP & roadmap
Shippable in about 8 weeks, then a path to production.
Build
- Two trades, three partner centres.
- Record a task, AI pre-score, assessor sign-off.
- Public verified passport page with QR code, in Arabic and English.
- Employer search with shortlist and contact request.
Not yet
- Fully automated scoring
- In-app employer payments
- More trades and countries
- Worker mobile app
Roadmap to production
Validate
Weeks 0–4- Interview 10 employers or their HR agents
- Sign 3 training centres
- Record and score 100 tasks by hand
MVP
Weeks 5–12- AI–assessor agreement above 85%
- 300 verified workers
- First 10 placements
Traction
Months 4–9- Employer subscription
- Re-training product
- Second destination country
Production
Months 10–18- 10 trades
- Outcome data product
- Seed round
09
Start this week
Your first steps. Tick them off as you go.
Metrics that prove it is working
- Verified assessments per month
- AI–assessor score agreement per trade
- Placement rate and salary of verified vs unverified workers
- Pass rate on destination-country skill tests
10
Sources
Every figure on this page links to its source. Facts reviewed 14 October 2026. Costs and market notes without a citation are our estimates.
- 1.ILO — International migrants are vital force in the global labour market (2024)
- 2.The Financial Express (RMMRU, BMET data) — Bangladesh's overseas labour migration up 12pc in 2025 (2026)
- 3.The Business Standard (BMET data) — Overseas employment up 11.27% in 2025 (2026)
- 4.The Business Standard (Ministry of Expatriates' Welfare) — Skill verification test results (2025)
- 5.UNB (Bangladesh Bank data) — Remittance inflows hit historic high of $35.56 billion in FY26 (2026)