When you’re running online courses with thousands of learners, even simple administrative tasks can balloon into weeks of work. Instructors, designers, and administrators all feel the pressure: repetitive processes eat away at time, and quality of learning often suffers as a result. I see this tension every day—between the scale of online education and the human effort required to sustain it. So I began looking for ways to bring automation and AI into our workflows, cutting down the routine and freeing people to focus on what matters most: teaching and learning. --- #### 1. Automating the Basics **⚠️ The problem:** One of our biggest time sinks was grouping learners. A single course might enroll thousands and we had to manually create and manage hundreds of groups in Canvas. If the course used Slack, we had to do the same thing again. Create channels and add learners one by one. Launching a course with high enrollment could take up to two weeks of effort. 💡 **What I did:** At first, I built a workaround using Google Sheets. I created templates that automatically generated group names and rosters, which cut down some of the repetitive work. It wasn’t perfect, but it helped us move faster before a long-term fix could materialize. In the meanwhile, I reached out to vendors from Canvas and Slack and eventually got this process fully automated. 🎯 **The result:** Groups and channels were created automatically, and learners were added instantly. What once took several hours was now done in minutes. --- #### 2. Making Media Searchable **⚠️ The problem:** Our design team had another bottleneck: images. Over the years, we had built a massive media library—but it wasn’t searchable. Many of the images did not have alt text or descriptive tags, which meant designers couldn’t easily find what already existed. They often ended up either scrolling the library for a long time before they stumbled upon the perfect image or recreating images from scratch. 💡 **What I did:** I built a program to loop through every image in the repository and run it through AI. For each image, the AI generated alt text, a detailed description (like “two people reading under a tree with books scattered around”), and even the HTML embed code. 🎯 **The result:** Suddenly, our library was searchable. Designers could reuse and embed images quickly, dramatically cutting down on duplicate work and speeding up course production. --- #### 3. Streamlining Course Development **⚠️ The problem:** Designing a course also has its share of repetitive tasks. Every course needs elements like learning objectives, rubrics, alt text for images, discussion questions, etc. These elements sometimes also need to be aligned with other standards - e.g. learning objectives to InTASC/ISTE. Designers were building these blocks again and again. 💡 **What I did:** To change this, I built several AI tools, like an alt-text generator, a learning objectives generator, a question generator, a rubric generator, etc. It could generate draft versions of these components on demand. Designers still reviewed and refined them, but they no longer had to start from scratch. 🎯 **The result:** These tools reduced the time it took to design courses, and gave our team more space to focus on the creative, higher-order parts of course design. --- > [!Tip] **The impact?** > Reduced deployment time for courses by 30% through automation, and improved design efficiency by 20% with AI support. > > By removing the drudgery from our work, we gave instructors and designers the time and space to focus on what really matters: connecting with learners and helping them learn.