Back to Work

Operations case study

Manager Binder

Operations tooling for shift-based teams, starting with a scheduling module that keeps constraints, review, and handoff visible.

ShiftDraft is the first concrete module in that direction: scheduling that makes constraints visible, keeps managers in control, and preserves the reasoning behind each draft.

What it is

An operations toolkit direction for managers who need continuity across recurring work, anchored by a concrete scheduling module.

Core problem

Manager work often leaks into spreadsheets, memory, message threads, and one-off rituals that only make sense to the person who last did them.

What I built

A template-driven scheduling workflow with deterministic draft generation, visible constraints, and a practical review handoff.

Why it matters

The product direction is not “automation replaces judgment.” It is operational memory: show the constraints, preserve the reasoning, and keep managers in control.

Scheduling module proof

The product is about manager control, not magic scheduling.

Templates define recurring coverage patterns instead of rebuilding the week from scratch

Availability, role fit, and hours expectations stay visible while drafts are generated

Validation checks and review flows make corrections part of the product, not an exception path

Exports and handoff states keep the schedule usable outside the app

Technical notes

  • Next.js App Router + TypeScript + Tailwind with Dexie/IndexedDB persistence
  • Multi-pass draft pipeline: warm start, assignment, must-fill swaps, hours balancing, and reassignment passes
  • Deterministic generation with validation checks and visible change trails during review
  • Operational UI focused on setup, generation, correction, and export

What it demonstrates

  • Constraint clarity matters more than flashy automation language.
  • Managers trust generated schedules when they can see and edit why decisions happened.
  • Browser-local tools can support serious workflow software when persistence and validation are first-class concerns.