Every learning team I’ve worked with has the same complaint: there’s never enough time. Projects are behind schedule, quality reviews get compressed, and the creative work that actually matters — designing great learning experiences — gets squeezed out by production overhead.
The instinct is to hire more people. But when I mapped the production workflow at Colibri, I found that roughly 60% of development time went to work that didn’t require human judgment: template setup, content assembly, formatting standardization, and publishing workflows. The team wasn’t slow. The process was bloated.
This is where automation changes everything — not by replacing the design process, but by eliminating the mechanical overhead around it. When I built TAGFORCE, the goal wasn’t to automate instructional design. It was to automate the pipeline so designers could actually focus on instructional design.
But here’s the part most automation advocates get wrong: not everything should be automated. QA review, pedagogical decisions, accessibility judgment calls, learner empathy — these require human expertise. The art is knowing the boundary. Automate the mechanical, preserve the judgment.
The framework I use: map every step in your workflow and ask two questions. Does this require context-dependent human judgment? Does the quality of this step depend on understanding the learner? If both answers are no, it’s a candidate for automation. If either answer is yes, it stays human — but you can still build tools that make the human faster.
The result at Colibri was a 60% reduction in development time and over $200K in annual savings. But the real win wasn’t efficiency. It was that the team started spending their time on work that mattered — and the quality of our courses improved because of it.