I built this because the system was failing my son.
It started with a Tuesday-night homework session. My 4th-grader, Michael, was on his 38th worksheet of subject-verb agreement. He got the first 35 right without thinking. He got the last 3 wrong — the same way he'd gotten them wrong all year. The system had spent six months teaching him what he already knew, and zero minutes teaching him what he didn't.
As an engineer, the diagnosis was obvious. Schools optimize for batch throughput, not individual mastery. Worksheets are cheap to assign and worthless to learn from. The 5 problems that would actually move him forward were buried under the 35 that wouldn't.
I built a tiny path engine: assess, find the gap, teach to the gap, retest, move on. In four months, he went from the bottom quartile of his class to reading two grade levels above. By the end of the year, three other parents at his school were asking for access. By the start of the next, eleven were.
I'm releasing this because what worked for one kid should work for thousands. I'm still the only one writing the path engine. Michael is still the QA team.