Course scheduling, under human control

Turn departmental rules and preferences into an editable, optimized schedule.

Configure the rules once. Import structured faculty and course inputs. Optimize teaching assignments and meeting times. Then use a local workbench to revise, checkpoint, compare, and export.

Tell me about your last cycle See the workflow

Please keep the first email high-level. Do not attach faculty records, preference files, or other confidential material.

✓ Deterministic optimization

✓ Local browser interface

✓ Reversible checkpoints

Synthetic walkthrough · fictional data

See the Schedule Workbench in action

Trouble with the player? Watch on YouTube

No black-box schedule

Rules and tradeoffs stay visible

No hosted scheduling account

A local browser-based operating model

No forced final answer

The department’s scheduling lead decides

From intake to export

A configured workflow, not a generic upload box.

Every department has its own canonical courses, rules, exceptions, and decision habits. The first cycle translates those realities into a tested configuration that can be reused.

Define the department’s rules

We map requirements, preferences, course categories, teaching loads, fixed commitments, and the decisions that must remain human.

Prepare a dependable intake

We configure an agreed Form or spreadsheet format and validate the inputs before they become a scheduling project.

Optimize instructors and times

Run deterministic proposals for teaching assignments and meeting times while respecting locks, availability, and hard constraints.

Review, revise, and export

Use the workbench to test changes, save checkpoints, compare versions, restore earlier work, and export the final schedule.

Built around judgment

Optimization that leaves you in charge.

The point is not to press a magic button. It is to make a difficult scheduling process faster, more consistent, and easier to reason about—without hiding the decisions that matter.

Explicit rules

Your approved rules are configured and testable—not inferred on the fly by a chatbot.

Deterministic optimization

Accepted inputs and configured rules produce a rule-driven proposal with visible locks and tradeoffs.

A real scheduling workbench

Optimization starts the conversation. You can lock, adjust, save checkpoints, compare, and restore.

Local browser interface

The pilot delivery design keeps the convenience of a browser while the scheduling application runs on the local device.

Department-specific setup

Every implementation begins by translating the department’s actual policies and intake into a tested profile.

Practical exports

Export clear spreadsheet and CSV outputs for review, downstream work, and institutional systems.

Early collaborator feedback

Less spreadsheet labor. Better working meetings.

An earlier scheduling implementation was developed in close collaboration with a large academic department. The department’s scheduling lead especially valued the editable workbench after optimization. This feedback describes that collaboration, not a claim of universal compatibility.

“I was skeptical that this was going to work, but I'm very enthusiastic now. Before, I spent 15–30 hours per semester working in our spreadsheets. Now, it's 1–2 hours, plus more efficient meetings with my colleagues.”
Department scheduler, University of Notre Dame · 40+ faculty

Where AI fits—and where it does not

The product’s scheduling logic is not generative AI.

AI-assisted development helps configure a department-specific profile and intake strategy. The scheduling application itself applies explicit rules and deterministic optimization.

That distinction keeps the rules and constraints inspectable and preserves a human-controlled workbench for every consequential decision.

Questions

The practical details.

Every department is a little different. These are the boundaries that stay consistent.

No generative AI participates in the scheduling logic. The scheduling application applies your department’s approved rules through deterministic optimization. AI-assisted development helps us configure each department’s profile and intake mapping, but the schedule itself is produced by explicit rules, optimization, and human decisions.

Direct implementation and support

Work with the person configuring the scheduling application.

I’m Graham Clay. I work directly with each department to understand its scheduling process, configure the rules and intake, test the resulting workflow, and support you through the first cycle.

Email Graham

A useful first conversation

Tell me how your last schedule was actually built.

We’ll discuss the inputs, rules, handoffs, and pain points from one real cycle at a high level and decide whether this is the right shape of problem for a pilot.

Start by email

Please do not email confidential records or attach faculty data. We will agree on a secure, limited data boundary before any files are shared.

automatedteach@gmail.com