Demonstration prototype — all candidate data is synthetic
CNY Talent Match

Method, evidence, and safeguards

A program-design prototype, not an automated hiring system.

CNY Talent Match shows how NYSDOL staff could turn upcoming demand into explainable candidate slates, training pathways, and measurable referrals.

10,000 synthetic records

OSOS-shaped work history, services, credentials, preferences, and deliberate data messiness.

Explainable matching

A fixed weighted score, human-readable factor contributions, and time-aware gap tiers.

Staff action

Outreach cohorts and named training routes remain recommendations for human review.

Data boundary

No OSOS data was accessed.

All candidate records are synthetic and were generated with a fixed seed. Aggregate county volumes are calibrated to public New York State LAUS unemployment counts; occupation mix to Central New York OEWS employment; education and veteran shares to American Community Survey county estimates. No names, records, exports, or credentials came from OSOS.

Deterministic scoring

Every point has a reason.

Occupation transferability30%
Required skills + semantic similarity30%
Credentials and education15%
Commute feasibility10%
Record recency / availability8%
Shift and wage preference7%

O*NET related occupations and alternate titles support transferability and missing-SOC normalization. Veterans and WIOA priority populations are never scoring inputs; priority changes order only inside equal five-point score bands.

Responsible use

The prototype uses deterministic, explainable factors and excludes demographic attributes from scoring. Referrals stay human-in-the-loop. A production service would run inside NYSDOL’s environment with role-based access, audit logs, data-retention controls, monitoring, and an appeal-and-correction workflow.

  • No automatic eligibility decision
  • No automatic rejection
  • No demographic scoring
  • Staff review before outreach

Path from demo to outcomes

Prototype → pilot → production

  1. 1Now

    Prototype

    Synthetic, statistically calibrated records demonstrate the workflow and make assumptions visible.

  2. 2Next step

    Pilot

    Run one real OSOS extract inside NYSDOL’s environment for one requisition wave and measure referral-to-hire conversion.

  3. 3After evidence

    Production

    Integrate governed data, role-based access, audit logging, monitoring, and staff feedback loops.

Central New York context

A long hiring horizon creates room to act.

Micron broke ground at White Pine Commerce Park in Clay on January 16, 2026. First concrete, originally planned for around September 2026, was reached ahead of schedule in July. Construction is expected to require thousands of skilled tradespeople; Fab 1 production is expected to start in 2030 as permanent hiring ramps.

The training ecosystem shown in gap recommendations includes Onondaga Community College’s $15 million Micron Cleanroom Simulation Lab, opened in October 2025; OCC’s Electromechanical Technology degree; and the Syracuse ON-RAMP advanced-manufacturing workforce center developed with CenterState CEO as part of New York’s $200 million ON-RAMP program.

Production work intentionally out of scope

  • Authentication and role-based access
  • Real OSOS or Micron integrations
  • Case notes or outbound messaging
  • Candidate self-service
  • Storage of real personally identifiable information
  • Fairness certification