Matt Murphy
Status: In development

Encore

I’m building one system for community-college continuing-education teams: catalog, registration and payment, class operations, attendance, completion, and certificates.

My role
Product owner and builder
Built with
Next.js, TypeScript, PostgreSQL on AWS with Drizzle, Clerk, Stripe Connect, Inngest, and Resend. Developed with Claude Code and Codex.
For
Continuing-education and workforce-training teams at community colleges
Status
In development
Policy centerVersioned, witheffective datesPolicy snapshotfrozen per enrollmentCatalogEach college’sown brandingRegister& payPriced atregistrationClassoperationsSessions, rostersand closeoutAttendanceTaken in theinstructor portalCompletionCertificates youcan verifyReportingEnrollment,attendance,revenuePayment goes straightto the college

The problem

Continuing-education learners are working adults who pay when they register and take short classes, often taught by industry instructors. Systems built for credit students model this poorly, so staff fill the gaps by hand.

Who it’s for

Continuing-education and workforce-training teams at community colleges: directors, registrars, operations and finance staff, instructors, and learners.

What I did

I researched the market, wrote the specification (more than a hundred planning documents, down to the state machines for enrollment, payment, class sections, and completion), designed the user interface and user experience, and decide what ships in every release.

Decisions that shaped it

  1. People shouldn’t be the integration

    Today the staff in CE operations and program development are the link between legacy systems, carrying information from one to the next by hand. Encore keeps the catalog, registration, payment, rosters, attendance, and completion in one system, so that joining is the software’s job.

  2. A platform, not a marketplace

    Each college keeps its own branded catalog. There’s no cross-college storefront and no take-rate.

  3. Policy frozen at enrollment

    Refund, attendance, eligibility, and pricing rules are versioned, and the rule in force is frozen onto each enrollment, so later changes never rewrite history.

  4. Price at registration

    Prices are calculated when someone registers, from policy, timing, and eligibility, rather than stored on the class.

  5. Operational depth over feature sprawl

    Deepen the screens staff already use instead of adding shallow new ones.

What’s built

  • Institution-branded catalog with search, cart, and time-limited seat holds
  • Registration and card payment, with refund workflows and a refund ledger (verified in test mode)
  • A versioned policy center with effective dates and audit reasons
  • A staff workspace that keeps a class’s sessions, instructor, rosters, and closeout in one place
  • An instructor portal for rosters, attendance, and grades
  • Completion rules, certificates, and a public verification page
  • Enrollment, attendance, and revenue reports

Constraints

  • FERPA. Learner data boundaries are designed around it: no education-record data in logs, analytics, or email subject lines.
  • Database isolation. Each college’s data stays separate, enforced in the application and in the database.
  • Role-based access. A permission matrix sets what each staff role can see and change.
  • Funds flow. Money moves directly to each college; the platform never sits in the flow of funds.
  • Accessibility. Screens are audited against WCAG, and the audits are kept.

Stuck with software that makes your staff do the integration?