Asobi

The description states that Asobi:

• Turns children's drawings into personalized learning experiences
• Uses OpenAI's vision capabilities to analyze drawings and extract structured information
• Generates educational lessons based on what was drawn
• Includes custom illustrations, narration, and questions connected to the original drawing
• Operates without requiring accounts, managing temporary lesson state in the browser
• Is built with Next.js, TypeScript, and uses OpenAI APIs for vision and text-to-speech

Evidence The author's own write-up.

Confidence Low — this is a self-reported product description with no independent verification of functionality or actual deployment.

The author states:

• The app starts with the child’s drawing rather than a pre-defined lesson
• It aims to make learning feel like a continuation of the child's own imagination
• Early versions struggled with making lessons feel connected to drawings
• Later versions focused on preserving visual identity and concentrating on one learning objective per lesson

Evidence Self-reported evolution in development approach.

Inference The positioning shifted from "AI-generated content" to "AI understanding creativity first, then building learning around it."

The description states:

• The target is children
• It focuses on children who draw before they can fully explain their ideas
• The experience is designed for personalization based on the child's own imagination

Evidence Author's own account.

Confidence Low — no evidence of customer segmentation, user research or actual users beyond the developer’s hypothesis.

The description states:

• No mention of pricing
• No indication of monetization strategy
• Lessons are generated without requiring an account
• Temporary lesson state is managed in-browser

Evidence Author's own write-up.

Confidence Not evidenced — no business model or pricing information provided.

The description states:

• Built with Next.js, TypeScript, React, Tailwind CSS
• Uses OpenAI APIs (vision and text-to-speech)
• Uses GitHub for version control
• Deployed on Vercel
• Includes Zod for schema validation
• Uses ESLint, Canvas, HTML5, Image processing

Evidence Author-declared tech stack.

Confidence Medium — the technical approach is described but not verified in terms of actual delivery or performance.

The description states:

• Submitted to OpenAI 2026 hackathon
• Team size: 1 person (Ukachi Benita)
• No mention of users, revenue, or adoption metrics
• No evidence of product-market fit or usage data

Evidence Self-reported project status.

Confidence Very low — no traction or maturity indicators beyond a hackathon submission.

The description states:

• No direct competitors mentioned
• The author claims to have approached the problem differently by starting with drawing instead of a lesson
• No evidence of market analysis or competitive landscape provided

Evidence Author's own positioning narrative.

Confidence Not evidenced — no competitive data, market size or competitor identification.

The description states:

• Single-person team (no evidence of scaling capability)
• Submitted to a hackathon — no indication of product development beyond prototype stage
• No revenue, customer or traction data
• Relies heavily on OpenAI APIs — potential dependency risk
• The author's own account suggests early-stage challenges in making lessons feel connected to drawings

Evidence Self-reported project details.

Confidence Medium — risks are inferred from lack of evidence and the prototype nature of the submission.

1. What is your definition of "personalized learning" in practice, and how do you validate that it works?
2. Have you conducted any user testing with children or parents? If so, what were the results?
3. How do you plan to scale beyond a single developer?
4. Are there any existing partnerships or pilot programs with schools or educational institutions?
5. What is your long-term vision for monetization and product development?