Classrooms where
teachers thrive.

It is one login and about ten minutes a day. Each playbook day is one complete practice: a short read, one action step to try in the room that day, and a reflection. The platform reads what teachers already log about their students. It never asks a teacher to write a report about themselves.

No. It is a classroom culture operating system. The 21 days start the habit. The product is the year-round loop: the Strategy Hub, the student log, walkthroughs, and a leader dashboard that shows who needs a move or a thank-you this week.

It produces aligned evidence of practice. It does not grade a teacher or a program, and it never predicts a rating. When a school turns on Texas T-TESS (K-8) or Texas Rising Star Category 2 (early childhood), each day shows the dimension it supports, and Day 21 produces an Evidence Summary your observers can use. Your observers still assign the rating.

Teachers never see the signal engine. A hard week reaches the principal as a coaching opportunity, and the teacher receives an assigned strategy from their leader, not an alert. A teacher's notes, reflections, and private journal are theirs: no leader can read them, and leaders see only that the work was done and how much was written. A teacher can choose to share one piece of their playbook writing with their principal, and stop sharing any time; journal entries can never be shared. What a teacher logs under What I've Tried (the strategies they tried and whether they helped) and pulse scores are visible to leaders. With T-TESS or TRS on, the Evidence tab shows which days have evidence and how much was written; a teacher who wants their evaluator to read a reflection downloads their evidence summary and shares it. We say all of that to staff plainly, because the trust only holds if the sentence is true.

Three features do, and none is ever sent a student's name. The coach's question box uses a language model (Claude, by Anthropic) to write a short answer from our own playbooks, strategies, and tools, and only from those. Names are removed on the teacher's device before a question leaves it. It never picks a consequence; that stays the school's call. A school can turn it off. A teacher's private journal can have the coach write back after an entry, but only if that teacher turns AI on for it; its suggestion is built from one day of our playbook and names that day, names are removed first, and no leader can read the journal or see anything from it. The platform owner has a question box over the platform's own counts, with school names swapped for codes before the model reads them. Everything else is the platform's own logic. No student ever signs in, and no AI ever reaches a student. Teachers log notes, incidents, and celebrations about their students; leaders see the patterns. When a school turns on the family loop, families open a tokenized link to a celebration sent home, with no login and no account. A teacher can also send a note home after trying a strategy, in family words, on the same kind of link; the family can say what helps at home, and that lands on the child's record for the adults who teach them and the school's leaders. No model writes those notes or reads the replies.

Texas T-TESS, Texas Rising Star, and Michigan Early Childhood are live today. A state’s rubric is data, not code: Tennessee TEAM, Florida FEAPs, California CSTP, and Danielson are the named next frameworks, and adding one is a seed file, not a rebuild. Districts get the District Lens: every campus on one screen, plays you can push to a principal, and a public demo.

Open the district demo