For workforce development organizations

Workforce data services: migration, integration, and custom software

Capstone Workforce ships the platform. Capstone data services build the pipeline that feeds it. State MIS migration, state wage-record integration, custom funder-format exports, and custom software builds against the same participant record and event stream the shipped modules run on. For organizations that need more than the shipped modules cover.

See the platform

The data wall every program hits

Compliance reporting was built for exit. Continuous improvement was not.

Most workforce programs run intake, coaching notes, and outcomes tracking across disconnected spreadsheets. Workforce Pell requires placement rates verified against state wage records, yet most institutions have no direct access to that administrative data and no established pipeline to obtain it. WIOA's exit-based measures were built for compliance reporting, not continuous improvement.

The accountability regime is arriving faster than the data systems needed to satisfy it. That is the gap the Data Quality Campaign has documented in every year of its policy work, and it is the gap Capstone data services are built to close.

The four service lines below map to the four recurring failure modes: source data scattered across systems, no pipeline to state wage records, funder formats the shipped modules do not cover, and program-specific needs no vendor ships out of the box.

Service lines

Four services against one participant record

Every service line writes to the same enrollment and event stream the shipped modules run on. Data services are additive to the platform, not a separate system.

Data migration and reconciliation

Load the state MIS export, the local compliance workbook a single person maintains, and every intake or placement spreadsheet the program runs on today into one participant record. Source-of-evidence tagged on every event. Column-mapping profiles saved so subsequent imports skip the manual step. Recompute-safe raw-import layer means a mapping fix retroactively corrects history without re-parsing source files.

State wage-record and longitudinal-system integration

For programs with a data-sharing agreement in place, configure the wage-record match against state SWIS/WRIS or a state longitudinal data system (Alabama P20W, Kentucky KCEWS, Washington ERDC, Texas ERC, Virginia LDS, and others). The buyer's SSN data never leaves the buyer's environment. Detailed mechanism below.

Custom funder-format exports

The platform ships PIRL-shaped CSVs and PDF board reports. For funders that require a proprietary layout — a state agency's own template, a philanthropic grantee report format, a workforce board's board-packet cadence — Capstone builds a one-off exporter mapped to the funder's schema and delivery window.

Custom software builds and analytics tuning

When the shipped modules do not cover a program-specific need — sector-specific interview rubrics, funder-partnership dashboards, historical cohort backfill so year-over-year comparisons work from day one, integration with a case management system already in place — Capstone builds against the platform's participant-record and event-stream contract.

How the wage-record match actually works

Four steps. Same mechanism whether you deploy Tier 3 or Tier 4.

State wage records are the primary evidence for WIOA Q2 and Q4 employment indicators. Most workforce providers have no direct pipeline to them. Here is the mechanism Capstone deploys.

Step 1

Deployment mode

The buyer's SSN data never leaves the buyer's environment. Two shapes: the Capstone container runs inside the buyer's AWS, Azure, or on-prem network (Tier 3 self-hosted), or the metadata-only connector runs even lighter and only computed indicator values sync back to the SaaS dashboard (Tier 4). Capstone ships updates via signed image pushes.

Step 2

Source ingestion

The container reads wage data from whichever pipeline the buyer already has. State-produced match file: the state UI or workforce agency runs the SSN match against its wage-record database and hands the buyer a matched-outcomes CSV, delivered via CSV, SFTP drop, or S3 bucket. Direct wage-record access: state workforce boards, community colleges with state DUAs, and public workforce agencies sometimes hold raw wage records under an active data-sharing agreement; the container runs the match locally. State longitudinal data system: the container consumes the state's endpoint into the buyer's participant record.

Step 3

Match execution

The wioaEngine wage-record match job (a standalone package deployed as part of the container) SHA-256 hashes the buyer's participant SSNs with a per-org salt. Only the participantHashId is stored; raw SSN is never persisted. The job compares against the state file's participant identifiers, either aligned hash or the state-assigned participant ID when the state runs its own match server-side. Successful matches materialize EmploymentEvent records with evidence.source set to wage-record-match and employer details when the state file includes them. Those events feed the WIOA Q2 and Q4 employment indicators automatically.

Step 4

Cadence and reconciliation

Weekly incremental. New extracts overwrite prior matches for the same participant-and-quarter pair because state data can revise as new UI records land, with full version history preserved in the per-org audit log so any change is visible to the buyer. Every read and write of a participant record is logged with actor, purpose, and timestamp, delivered to the buyer as a weekly S3 export they control.

What Capstone delivers

  • The self-hosted container or metadata-only connector, built from the same codebase as the SaaS platform
  • The wioaEngine wage-record match job, published as a versioned standalone package
  • Ingestion adapters for CSV, SFTP, S3, and REST API depending on how the state delivers
  • State-specific schema configuration (each state's PIRL-matched output has slight column-name variations)
  • Verification against known-good test cases before the first live run
  • Ongoing updates and state-schema drift maintenance

What the buyer's IT does

  • Deploy the container into the buyer's AWS, Azure, or on-prem environment (one-time)
  • Hold the SSN-to-hash mapping locally
  • Establish and maintain the data-sharing agreement with the state UI or workforce agency
  • Route the state file into a location the container can read (S3 bucket, SFTP endpoint, database connection)
  • Monitor the container's health via the standard container-orchestration tools the buyer already uses

Where the data lives

Tier 3 or Tier 4. Your SSNs never leave your environment.

Wage-record integration runs under one of two data-sovereignty deployment shapes. Both use the same match mechanism.

Tier 3 — Self-hosted container

The Capstone container runs inside your AWS, Azure, or on-prem environment. Your participant data, your SSN mapping, and your wage-record files never cross your network boundary. Capstone ships container updates via signed image pushes. Your IT deploys once and monitors with the same tooling you already use. Add-on to the SaaS platform for organizations that require on-premise or private-cloud deployment.

Tier 4 — Metadata-only connector

The lightest footprint. A buyer-side connector runs the calculation engine locally against your data and pushes only computed indicator values and intervention-queue metadata (no PII) to the Capstone SaaS dashboard for rendering. For maximum-restriction buyers where participant data must remain entirely within the buyer's network. Same match mechanism, minimal outbound surface.

Pricing

Three tiers, self-qualify before you book a call

Bands reflect real deal-sizing variance in the workforce vertical. Use the pricing calculator for a scoped estimate against your specific situation.

Starter

$25K – $60K

one-time

Single dashboard, outcomes verification, data cleanup, or a single-source migration with one program year of history. The on-ramp tier for grant-constrained nonprofits and smaller reentry programs.

  • Data cleanup or single-source migration
  • One dashboard or outcomes surface
  • Deployment on the SaaS platform

Core

$75K – $200K

one-time

Custom case management or outcomes system plus 1–2 integrations. Multi-source migration with 2–3 years of backfill. Sector-specific rubrics or funder-partnership dashboards. The sweet spot for LWDBs, community colleges, and mid-scale nonprofits.

  • Multi-source migration + backfill
  • Single state wage-record integration (SWIS/WRIS or P20W)
  • 1–2 custom funder-format exports
  • Custom module extension

Enterprise

$200K – $400K+

one-time

Multi-organization or multi-state builds with several integrations and full historical migration. State-agency deployments under Tier 2 GovCloud or Tier 3 self-hosted. The tier that competes with legacy $1M+ workforce system replacements at a fraction of the ceiling.

  • Complex multi-source migration
  • Multi-state wage-record integration
  • 3+ custom funder-format exports
  • New module or multi-module custom build
  • GovCloud or self-hosted deployment

Estimate your specific scope

Model annual license + one-time services against your program size

The interactive calculator returns an annual recurring band, a one-time services band, a 3-year total, and a grant-math anchor that back-solves the grant size supporting the estimate as a 5–10% tech line under 20 CFR 683.205.

Open the calculator

Frequently asked questions

Direct answers about data services engagements

Which states can Capstone integrate wage records against?

The mechanism works against any state that provides either a matched-outcomes file (most common), direct wage-record access under a DUA, or a state longitudinal data system endpoint. We have working configurations against Alabama's P20W, and reference architectures for Kentucky KCEWS, Washington ERDC, Texas ERC, Virginia LDS, and the standard SWIS/WRIS interfaces most states expose. New state schemas are ~2-3 days of configuration work.

Do you handle the data-sharing agreement with the state, or does our IT?

The buyer's program office holds the DUA with the state. That agreement is a program-to-agency relationship and can't be delegated to a software vendor. Capstone provides the technical documentation the state IT reviewer will ask for (data-flow diagrams, encryption-at-rest and in-transit posture, per-org audit log, hash-based PII minimization), and joins the review calls as needed.

How long does a data migration typically take?

Scoped per engagement. A single-source migration (state MIS export or a local workbook) with one program year of history typically completes in 2-3 weeks including validation. Multi-source with 3+ years of historical backfill runs 6-10 weeks. The critical path is usually the buyer's side (assembling the source files and reviewing the column-mapping profile), not the Capstone side.

Can you build a custom funder-format export we do not have a template for?

Yes. Send us the funder's current template (PDF, CSV, XLSX, or a written spec), the delivery cadence, and a sample of the current manual output. Turnaround is typically 3-4 weeks for a first version, with iteration against the funder's actual reviewer as the tuning cycle.

What is the pricing model for data services?

Three tiers. Starter: $25K–$60K one-time for a single dashboard, outcomes verification, or data-cleanup engagement. Core: $75K–$200K for custom case management or outcomes systems bundled with 1–2 integrations. Enterprise: $200K–$400K+ for multi-org or multi-state builds with several integrations and full migration. Sovereignty add-ons run $25K–$150K per year depending on tier. Use the pricing calculator at /workforce-pricing-calculator to model your specific scope, or bring a scoping call for a written proposal within a week.

Can data services run without the full Capstone platform?

The data services are designed to feed into the Capstone platform. A migration destination other than Capstone, or a standalone custom-export engagement, we can look at case-by-case. State wage-record integration and continuous compliance calculation require the container or SaaS to consume the outputs, so those services do not run standalone.

Start with a scoping call

30 minutes with the engineering team, not a sales pitch

Bring your current data footprint (state MIS, funder templates, any spreadsheets that hold placement or credential evidence today) and the reporting deadlines you have coming up. You get a scoped proposal within a week.

See the platform