Doledesk: Automating Substitute Teacher Scheduling

In 2023, I started looking into inefficiencies in school operations - areas where automation could replace tedious, repetitive processes. Automation was becoming increasingly practical, and I wanted to build something with real impact. That led me to a conversation with my high school's vice principal. I asked a simple question: “What are the most time-consuming tasks you deal with daily?”

The answer was clear: substitute teacher scheduling. Every morning, administrators scrambled to check which teachers were absent, cross-referencing a massive spreadsheet to manually assign substitutes. It was an inefficient, error-prone process - one that was ripe for automation. That insight led me to build Doledesk, a system designed to fully automate substitute teacher scheduling using a rules-based algorithm.

The Complexity of Scheduling

At first, scheduling substitutes seemed like a straightforward problem - identify absent teachers and match them with available subs. But the deeper I went, the more complexity I uncovered. Several constraints made this a non-trivial problem:

• Legal Compliance: Many school districts have strict labor laws. For example, a teacher can't work more than three consecutive periods without a break.

• Subject Matching: Not all substitutes can teach every subject. A math teacher shouldn't be assigned to an English class.

• Multi-Layered Dependencies: If one substitute isn't available, the entire schedule may need to be reshuffled dynamically.

• Data Privacy Restrictions: Storing teacher data required careful adherence to K-12 privacy regulations, meaning traditional database solutions weren't viable.

These constraints required a robust, flexible system capable of handling real-world edge cases while optimizing for efficiency.

Building Doledesk

The core of Doledesk was a backend system built with Java and JavaScript, designed to process teacher absences and dynamically assign substitutes. Here's how it worked:

• Data Input Pipeline: Each morning, a fresh list of absent teachers was fed into the system. Since storing persistent data wasn't an option due to privacy regulations, all scheduling had to happen in real-time.

• Algorithmic Matching: The backend used a constraint-satisfaction algorithm to assign substitutes based on availability, subject expertise, and legal guidelines. If an optimal match wasn't found, the algorithm recursively adjusted placements.

• Automated Notifications: Once schedules were finalized, the system sent out automated email notifications to substitutes and teachers.

• Failsafe Mechanisms: If any substitute declined their assignment, the system reran the matching algorithm to fill gaps dynamically.

For the UI, I opted for Bubble, a no-code editor, to accelerate frontend development. This allowed me to quickly iterate on the UI and get a working product.

Deployment and Impact

Once Doledesk was live, my school used it for a month. The result? Administrators who previously spent hours manually assigning substitutes were now completing the process in minutes. What used to be a chaotic, last-minute scramble was now a structured, automated workflow.

While this project started as an experiment, it quickly became something more - proof that automation can significantly reduce administrative overhead in education. Doledesk wasn't just about saving time; it was about ensuring that students always had the right teachers in place, improving the overall classroom experience.

What's Next?

Doledesk validated an important idea: many outdated, manual processes in education can be automated with the right approach. Looking forward, I see opportunities to expand this concept beyond substitute scheduling - perhaps into broader school operations or even district-wide automation tools.