video1.0<iframe src="https://www.loom.com/embed/be31d8fb27ee4b71a4cfb95c9ad9a4bb" frameborder="0" width="1670" height="1252" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>12521670Loomhttps://www.loom.com12521670https://cdn.loom.com/sessions/thumbnails/be31d8fb27ee4b71a4cfb95c9ad9a4bb-3c2554a60a3c9951.gif389.823This Loom presents a hackathon submission focused on analytics for class and group retention, plus an AI question tool for staff. The author uses a multi year dataset (about 10 years, fall and spring since around 2018) stored in SQLite, with analytics driven by SQL queries (no AI used for analytics) and generative AI used to formulate dashboard text. They review outcomes for the adults in twenties classes, including a spring 2020 comparison showing one class as atypical, and they show group performance by identifying groups losing new people, multiplying, and splitting off (including a Cree group split with eight people). The second half demonstrates an AI powered question interface built on Next.js with TypeScript and Tailwind, where prompts query the SQL database and return retention metrics such as keeping eight out of 23 versus a typical 59%, with 10 out of 26 keeping the same group. They also recap that the biggest group fall off in each year occurs in the first year and highlight first year signups from fall 2025, including August 25 left dates.