OpenAI 2026 hackathon

Wizard Math

Adding an AI companion layer to Wizard Math, an existing math adventure iOS game I'm building. Each pet has a Context-Engineered personality players can bond with.

Solo project by Harry Zhu · 0 likes · 0 comments

Archive position — measured, not model output

0 likes on Devpost

2,264 of the 7,856 archived projects have more likes, and 5,592 share exactly 0 — so this project's #7,710 place in the like-ranked listing is a tie-break inside that group, not a ranking.

Projects (log scale)

1
10
100
1k
10k
05,592
11,758
2285
3–4132
5–975
10+14

Likes on Devpost. ▲ marks this project's group.

Show the figures
LikesProjectsShare of archive
05,59271.2%
11,75822.4%
22853.6%
3–41321.7%
5–9751.0%
10+140.2%
Devpost like counts for all 7,856 archived projects, captured when this archive was built.

Executive Summary

What the company appears to be

Wizard Math is an iOS game that the author, Harry Zhu, describes as a math adventure game where players solve math problems to fight monsters, earn coins, and raise pets. The project adds an AI companion layer to the existing game using OpenAI's API, with each pet having a context-engineered personality shaped by mood, health, and happiness. The AI chat feature is implemented via a proxy service that manages API keys, rate limits, and contextual grounding for safety.

What changed

The author states that prior to this project, the pet system in Wizard Math was one-sided — pets reacted but did not talk or remember events. This update introduces an AI-powered chat layer where pets can engage with players emotionally and contextually, enhancing player bonding and engagement.

Single most important open question

Is there a viable business model for sustaining long-term AI costs associated with the pet companion feature? The description indicates that the current approach uses a shared OpenAI account and lacks any revenue or pricing mechanism to cover ongoing usage.

Back to contents

What The Product Actually Is

The description states:

  • Wizard Math is an iOS game involving math problem-solving, monster battles, coin earning, and pet adoption.
  • It includes a pet system where pets can be fed, walked, and assist in battles.
  • A new AI chat layer was added using OpenAI’s API (specifically gpt-4o-mini).
  • Each pet has a personality influenced by its mood, health, and happiness.
  • The pet remembers recent events and conversations.
  • If the AI service is unreachable, it falls back to offline replies.

This is an iOS game with an AI-enhanced companion layer. It is not yet released on the App Store.

Back to contents

Positioning & Claim Evolution

The description states:

  • The goal was to make math practice feel like adventuring with a friend.
  • Prior to this update, pets were reactive but lacked emotional connection or dialogue.
  • The author claims that AI chat makes the pet "feel like it knows the game" and encourages players to keep going.
  • The AI companion is described as more than just a scripted response — it's meant to act as a lightweight tutor and emotional bond.

Positioning: The product positions itself as an emotionally engaging math learning tool for children or casual gamers, enhanced by AI companionship. It claims to improve retention and engagement through personalized interaction.

Inference: The author implies that this is a novel approach in gaming — using AI to create flexible, dynamic NPCs with personality and memory. However, no evidence of prior similar products or market differentiation is provided.

Back to contents

Target Customer & ICP

The description states:

  • The game targets players who enjoy math adventures and pet care.
  • The author’s daughter was a key user during development.
  • It's described as a casual game that could appeal to families or children interested in both math and fantasy themes.

Not evidenced: No explicit customer segments, demographics, or ICP defined beyond "casual gamers" or "families."

Inference: Based on the narrative, the target may be young players aged 6–12, or parents looking for educational games. However, this is inferred from the author’s personal experience and not substantiated.

Back to contents

Business Model & Pricing Evidence

The description states:

  • The game has not yet been released on the App Store.
  • It uses a shared OpenAI account for API access.
  • There is no mention of in-app purchases or subscription models.
  • The author notes that a one-off purchase won’t cover long-term AI costs.
  • Cost estimates are made for 1k players chatting simultaneously for a month (~$50/month).
  • A business model has not been determined yet.

Not evidenced: No pricing, monetization strategy, revenue streams, or customer acquisition costs mentioned.

Inference: The author seems to be exploring whether the game can support AI features through direct sales or if additional monetization (e.g., premium pets, chat upgrades) is needed. This remains an open question.

Back to contents

Technical & Delivery Signals

The description states:

  • Built using Xcode and Swift.
  • Uses Codex and GPT-5.6-SOL for development assistance.
  • The AI chat feature uses gpt-4o-mini due to cost considerations.
  • A proxy service handles API keys, rate limiting, prompt injection protection, and context management.
  • The proxy combines game rule catalogs with pet state for grounding AI responses.
  • UI design was guided by AI to match the game’s dark-fantasy identity.
  • The author used spec-driven development and iterative feedback loops with GPT.

Signals:

  • Strong technical integration of AI into a mobile app.
  • Use of proxy architecture for security and control.
  • Context-awareness built into the AI pipeline.
  • Spec-driven development approach shows planning and clarity in execution.

Inference: The use of AI tools like Codex and GPT 5.6 suggests that the author is leveraging LLMs extensively for both design and code generation, which may indicate a scalable development process.

Back to contents

Traction & Maturity Signals

The description states:

  • The game has not yet been released on the App Store.
  • It was submitted to an OpenAI hackathon.
  • The prototype proved the experience.
  • No mention of user testing, downloads, or usage metrics.

Not evidenced: No evidence of traction, customer base, or product-market fit beyond internal testing and a hackathon submission.

Inference: The project is in early development phase. It has functional proof-of-concept but lacks real-world validation or market entry.

Back to contents

Competitive Context

The description states:

  • No mention of competitors.
  • The author does not reference other math games or AI-powered pet companions.
  • The idea of integrating AI into pet/NPC systems appears novel within the context described.

Not evidenced: No competitive analysis, market size, or existing players in the space.

Inference: If the concept is unique, it may face low competition. However, without a clear understanding of the broader market, this cannot be confirmed.

Back to contents

Key Risks & Red Flags

The description states:

  • The business model for AI costs is unresolved.
  • There is no clear path to monetization or revenue generation.
  • Prompt injection and hallucinations are risks that were mitigated via proxy design.
  • Player messages can contain malicious prompts, which must be filtered by the proxy.

Red flags:

  • Lack of financial sustainability — reliance on shared OpenAI account without a scalable model.
  • No evidence of user testing or feedback loops beyond author’s personal experience.
  • Risk of AI hallucinations if context is not properly managed.
  • Dependency on one developer (team size = 1), which raises scalability concerns.

Inference: The project may be technically feasible but lacks commercial viability or long-term strategy.

Back to contents

Diligence Questions To Ask The Founders

  1. What is the current plan for monetization and how will AI costs be covered?
  2. How does the proxy service handle data privacy and compliance (e.g., GDPR)?
  3. Has there been any external user testing beyond the author’s daughter?
  4. Are there plans to expand beyond iOS or add more AI features?
  5. What are the key assumptions behind the AI integration, and how were they validated?
  6. How does the team intend to scale beyond a single developer?

Back to contents

Investment/Partnership Verdict

The description states:

  • The project is in early prototype stage.
  • It has not yet been released or monetized.
  • The author is exploring AI integration as part of a larger game development journey.

Not evidenced: No financials, revenue projections, or traction data available.

Verdict:

This is an experimental, self-developed prototype with strong technical execution and emotional engagement claims. However, it lacks commercial viability indicators such as monetization plans, user feedback, or market validation. The author’s personal investment and AI tooling are impressive, but the project does not yet demonstrate a scalable business model or clear path to revenue.

Confidence: Low — Based on self-reported evidence only, with no external verification or traction data.

Back to contents

Source

Submitted to the OpenAI 2026 hackathon on Devpost. Project home on DevPost.

The analysis above was generated by a language model from the project's own one-line description. It is not independent research and contains no verified traction, revenue or customer data.