The vision

<aside> 💡

Help teachers generate high-quality, standards-aligned quizzes instantly with the help of AI, without sacrificing control, accuracy, or cost-efficiency.

</aside>

Problem Space

Teachers in the U.S., our primary user segment, often:

• Struggle with time constraints while creating assessments
• Need content aligned with state-level standards (e.g., Common Core, NGSS, TEKS)
• Want trusted, verifiable sources for learning content
• Are skeptical of AI-generated content, especially when it hallucinates or gives irrelevant results

Company constraints

• GPT-4 and other AI LLM tools were costly at scale, every bad request burned tokens
• Needed to make AI “useful” without leaving a blank text box for teachers to fill (which often led to junk output)

Design Goals

1. Help teachers “prompt less, guide more”
2. Localize results to their subject, grade, and standards
3. Prevent AI hallucinations and surface verifiable content
4. Use credits wisely, reduce retries and spam queries

🔍 Discovery & Research

We began with: