The shape of the case is present; the evidence for it is not yet.
60 Developing · 78 Promising · 96 Strong
The weighted total, computed in code from the evidence ratings for this deck. The same file scores the same every time: the ratings are made once and reused, so re-uploading returns this number, not a near one.
Traction (8.5 of 10) is the strongest supported category, with live deployment at the anchor university across 14 professors and 5 courses, 4,000+ students assessed, 92.4% satisfaction, and ₩70M (~$50K) in booked revenue described as a paid contract. Problem (6.8 of 9) is backed by a field study of 272 professors, TAs and students across 9 universities, finding 20+ hours to grade one 300-student assignment, 13 hrs/week per TA, 26.3% score variance between graders, and 72% of students receiving no personalized feedback. Value Proposition (4.1 of 6) draws on the same deployment: 3.8x faster grading, 96.8% agreement with professor scores across 13,800+ verified items, variance cut to 11%, and an 80% drop in grade appeals.
Why Now (3.2 of 7) rests on an assertion that multimodal AI became viable only in the last 18 months, but no dated external source is given for this or for the claimed shift away from unsupervised take-home essays. Market Opportunity (6.2 of 12) cites a $3.7B language-assessment market and 40,786 universities running 240,000+ English courses, but neither figure is sourced. Team (8.8 of 17) names two founders from the same university with no U.S. team members or advisors, and Defensibility (4.6 of 8) rests solely on a 13,800-item verified grading dataset with no patents or exclusivity claimed. Financials (5 of 9) show booked and projected revenue (₩70M actual; ₩80M and ₩200M+ projected for 2026 and 2027) but no unit economics, burn rate, or runway figures. Overall this reads as a Seed-stage deck for the Korea-based case but a pre-seed case for the U.S. market, where Validation/Traction sits at 1.2 of 15 and Pricing & Business Model Fit at 1 of 10.
The deck invites the question of how the anchor-university traction — all Korea-based revenue, users, and validation — translates into the ~19M-student U.S. market it cites, given zero named U.S. pilots, customers, or team members. It also invites a question about unit economics behind the projected 200%+ YoY revenue growth, since no burn rate, runway, or per-institution cost figures are provided beyond the stated $14K (₩20M) per-institution licence price. Answering either would require named U.S. institutional engagement alongside a stated customer-acquisition cost.
Colour reads how well the category is evidenced, judged the same way at every stage. Track length reads what the category is worth at Seed — Team at 17 points is the longest, Ask & Use of Funds at 5 the shortest. So a bar can be fully green and still contribute little, or red and cost a lot. Vertical rules mark where Traction and Go-to-Market cross into Moderate and High risk at Seed.
Significant customer, competitive, GTM, or execution validation is still required.