Show HN: I'm a teacher and built an AI presentation tool

Target users and use cases

• Designed mainly for K–12 teachers, with an age selector from kindergarten to adult; users confirm it can fit primary, secondary, and even introductory tertiary courses.
• Several see strong potential beyond schools: corporate training (e.g., fine foods, internal procedures with RAG on company docs), ESL teaching, and employee onboarding.
• Many say its biggest value is for quick prep: last‑minute lessons, worksheets, quizzes, and plenary activities rather than full, polished courses.

Teaching quality, accuracy, and AI “slop” concerns

• Critics report verbose, generic text, shallow treatment of topics, confusing quiz wording, and at least one outright wrong answer (MP3 vs WAV example).
• There are worries that weak teachers will present AI output unedited, normalizing low‑quality, error‑prone materials and undermining genuine pedagogy.
• Others counter that teachers are professionals who already curate textbooks, videos, and worksheets, and will review and edit AI content accordingly; AI is framed as a “starting point” and force multiplier, not a replacement.
• Some note that pre‑AI materials were often poor too; AI may be no worse than existing low‑effort resources if used thoughtfully.

Slide design and UX

• Common feedback: slides are too text‑heavy; better for classroom handouts than conference‑style talks.
• Discussion around “walls of text” vs richer speaker notes; several suggest short bullets plus detailed notes.
• Themes and customization exist but aren’t always obvious; image choices can be off, though swap controls help.
• Non‑English output is inconsistent; users ask for better multilingual support.
• Slides can be exported to HTML for local viewing but only edited in‑app, raising concerns about long‑term reliance on a solo‑dev service.

Teacher workload and curriculum context

• Multiple educators stress how much time is spent constantly revising materials: adapting to new cohorts, changing priorities, personalized learning plans, and frequent subject reassignments.
• Others from more textbook‑centric systems are surprised at this and question what needs revising; discussion highlights pacing, sequencing, differentiation, and local curriculum shifts.
• Some debate the broader question of slides vs blackboard and of live “performance” teaching vs self‑directed reading and exercises.

Tech stack, business, and trust

• Built with PHP backend, vanilla JS/jQuery frontend, Node + socket.io for ChatGPT interactions; token costs currently absorbed by the creator.
• Pricing and free‑vs‑pro feature messaging are seen as a bit confusing.
• Strong calls for clear privacy policy and TOS, and for explicit limits on entering any student PII, given the education context and solo‑operator risks.

Competition, differentiation, and roadmap

• Commenters note many competing AI‑for‑teachers tools; they suggest focusing on the interactive activities (quizzes, word searches, cloze, crosswords) as the key differentiator.
• Ideas raised: LMS integrations to ease procurement, a marketplace for refined slide decks, collaborative features for students, and better animation/visualization support.