SPG Research · AI Assurance Portfolio
WorkMatch
Better information, better matches.
A skills-first, evidence-weighted employment matching engine that improves the quality of information used in hiring decisions — without deciding who gets hired.
The problem
A poor first match costs everyone — and compounds.
- The candidate loses the job, and their confidence to try again.
- The employer concludes inclusive hiring "doesn't work" and resists next time.
- The agency loses credibility and spends its time repairing failed placements.
The failure is upstream: the information used to make the match is thin, inconsistent, and not trusted.
The insight
Fix the information, not the decision.
WorkMatch improves match intelligence before placement — structured skills evidence, requirement-validated jobs, and explainable briefs. It is decision support, not automated hiring. That distinction is the whole product — and its legal moat.
What it is
One guided intake → many trusted outputs.
- Skills Passport — evidence-weighted, claimed→validated, bilingual.
- Job Demand Profile — requirements separated, employer-validated.
- Match Confidence Brief — a qualitative band with reasons, never a score or ranking.
- Résumé + Skills Development Plan — from the same intake, no re-entry.
The differentiator
Self-directed first. Assistance optional.
For people
A candidate can build a profile and become match-ready with no agency required — including newcomers, older workers, and the general population, not only supported employment.
For agencies
Staff add value where it counts instead of becoming a mandatory bottleneck. Agency support is an enhancement layer, not a dependency — which is what makes the market scalable.
Governance is the moat
Compliance built in — and provable.
- No automated rejection · no ranking · no hire/no-hire score.
- Client-owned, consent-based, human-reviewed; funders see aggregate only.
- Aligned to Quebec Law 25, Ontario's AI-disclosure rule, and Canadian privacy expectations.
- The guarantees are enforced as tests and database constraints — not promises.
In a market where AI-hiring tools face rising legal risk, "provably safe" is a durable advantage.
It's real, today
Not a concept — a built, clickable system.
- Live interactive demo (candidate → agency → employer → funder on one dataset).
- Clickable self-directed prototype + full workflow wireframes, bilingual.
- Validated data contracts, a scaffolded repo, and the trust-spine coded.
See it: workmatch.spgresearch.com → Live demo.
Market & buyers
Employment infrastructure, not a job board.
Who pays
Provinces/funders (anchor licence) · agencies (SaaS) · employers (light, later). Standardized, funder-ready outcome reporting.
Portfolio fit
WorkMatch is the Screening/WorkMatch engine in SPG's AI Assurance Portfolio — it feeds qualification results into the broader assurance stack, opening a second route to revenue.
Status
Specification-complete; ready to build & pilot.
- 12 execution-grade sprint runbooks; MCS gates specified as tests.
- Data schemas validated; trust-spine (tenancy/RBAC/audit/consent) coded.
- Deployable engine site + demo; pilot toolkit + warm agency shortlist ready.
Bottleneck is execution & validation — not strategy.
Business model
Anchor on funders; land via agencies.
- Provincial / funder licensing — the anchor: standardized outcomes across a network.
- Agency SaaS — the entry: per-agency tiers, bilingual, funder-ready.
- Implementation services — onboarding, bilingual validation, migration.
- Land with a no-cost 12-week pilot; convert on evidence.
The ask
One anchor pilot. One believer.
We're securing a first agency + funder pilot and the resources to build the Minimum Credible System. Everything to start is already written, validated, and demonstrable.
Let's improve the match — before the placement.
SPG Research · workmatch.spgresearch.com · [your contact]