Pilots
One gateway course. One term. A real number.
We are recruiting two or three design partners to run one course through the recovery loop for six to eight weeks, measure it with their research office, and tell us where it breaks.
One gateway course · one term
8 weeks
- Set upWeek 0
- Launch from your LMS
- Roster imported
- Instructor publishes material
- RunWeeks 1–6
- Students practise and ask Ava
- Retries and hand-offs recorded
- Weekly instructor debrief
- AdjustMid-point
- What to reteach, by concept
- Gaps in the material
- Send a retry to who is stuck
- ReportWeek 8
- Recovery measures
- Course outcome vs baseline
- Reviewed with your IR office
Five tracks
Pick the course that fits your institution.
Each track is a course type the product runs end to end today. A track we do not list is one we would decline.
Community college gateway courses
College Algebra, gateway science and English — the courses the retention office already tracks.
The course that already has an instructor, a syllabus and a completion problem. Faculty publish; students practise and ask Ava; the instructor sees who is stuck and what to reteach before the exam.
Shipped today: The loop runs end to end on College Algebra, cited against the instructor’s own material. Misconceptions are named where the course carries faculty-signed wrong-answer meanings.

Pilot terms
Written like a term sheet, not a brochure.
What we ship
- The full product, configured for your institution.
- A weekly instructor debrief for the whole pilot.
- A live cohort view for the section.
- An evidence report your research office can share.
- No cost during the design-partner period.
What we ask for
- One course, one section, one instructor.
- Roughly 20–40 students for at least six weeks.
- A named contact in institutional research.
- Permission to write it up; you review first.
Not included
- Accreditation or exam eligibility.
- Curriculum, instructor training or lab supervision.
- A promise that any number moves. If it does not, that is the result.
What a pilot reports
Recovery, not chatbot usage.
Defined with you before the pilot starts and reviewed with your research office. We have not run a full-term pilot yet; these are the measures the first one exists to produce.
- First-retry recovery
- Share of struggles resolved on the student’s next attempt.
- Eventual recovery
- Share resolved within the term, and how long it took.
- Repeat misconception
- Students who made the same mistake again after help.
- Hand-offs to people
- How many reached a tutor or instructor, and what happened next.
- Material gaps
- Questions your material could not answer, by topic.
- Course outcome
- Assessment scores and drop / fail / withdraw rate against your baseline.
01
02
03
04
05
06
Two paragraphs is enough.
The course, the section size and one number the institution already tracks. We reply with a plain answer about fit.