MyTurn

The description states that MyTurn allows users to describe a real person from their life and then engage in a spoken conversation with an AI version of that person. It does not grade the user but instead shows how the story shaped, what phrases worked, and provides a saved "story card." The AI is described as being trained to respond naturally, with repair mechanisms rather than interrogation.

• Product function: To simulate conversational interaction with AI characters based on real-life relationships.
• User interaction model: Spoken conversation via browser-based speech API; feedback provided in text form.
• Technology stack: Next.js, TypeScript, Tailwind, deployed on Vercel; GPT-5.6 for conversations and feedback; Codex for development.

Note: The description does not state whether the AI character is customizable or if users can edit or refine their story cards after a conversation.

The author positions MyTurn as a tool for language learners who struggle with speaking due to perfectionism. It claims to help users overcome hesitation by allowing them to practice in a safe, non-judgmental environment.

• Core claim: "Stop waiting and start speaking" — encouraging participation over perfection.
• Evolution of positioning: The project started as a personal solution for the author’s own language learning challenges and evolved into a tool that could help others with similar issues.
• Narrative shift: From individual improvement to building a system where people can return to conversations with AI versions of real individuals.

Inference: The product may be positioned more broadly in the future as a conversational AI platform for storytelling, education, or therapy — but this is not yet evidenced.

The description states that MyTurn was built for someone who studies how we learn language and experiences hesitation in group conversations. It targets individuals who want to improve their speaking fluency through practice.

• Primary persona: Language learners (especially those with low confidence or perfectionist tendencies).
• Secondary use case: People looking to rehearse difficult conversations (e.g., interviews, exams).

Not evidenced: No data on specific demographics, geographic reach, or user segmentation.

There is no evidence of pricing structure, monetization strategy, or business model in the description.

• Monetization: Not stated.
• Pricing: Not stated.
• Revenue streams: Not stated.

Inference: If this were to scale, it might involve subscription models or premium features for advanced storytelling or character customization — but no such evidence exists.

The project was built using modern web technologies and AI APIs:

• Frontend: Next.js, TypeScript, Tailwind
• Backend/Deployment: Vercel
• AI Tools Used: GPT-5.6 via Responses API; Web Speech API for speech input/output
• Development Tooling: Codex used in a workflow where the author directed product decisions and Codex wrote code

Not evidenced: No information on scalability, infrastructure, or long-term technical architecture.

There is no evidence of traction, adoption, or user engagement beyond the author’s own experience.

• User base: Not stated.
• Engagement metrics: Not stated.
• Retention: Not stated.
• Product maturity: The project appears to be a hackathon submission with limited production history.

Absence of evidence: No data on how many users have interacted with the system, how long they spent using it, or whether they returned.

The description does not mention competitors or similar products. It is unclear if there are existing tools in this space that offer AI-driven conversational practice or character-based storytelling.

• Direct competitors: Not mentioned.
• Indirect competitors: Possibly language learning apps, chatbots for therapy or education, or generative AI platforms for roleplay.
• Market positioning: Unclear due to lack of competitive analysis.

Inference: If the product gains traction, it could compete with tools like Duolingo, Grammarly, or AI roleplay platforms — but this is speculative.

Several risks and red flags emerge from the self-reported nature of the description:

• No revenue or monetization strategy: The project appears to be experimental.
• Single-person team: Limited capacity for scaling or iterating quickly.
• No user feedback loop: No indication that users have provided input or engaged with the system beyond the author’s own use.
• Unverified AI behavior: The description says the AI avoids inventing details, but there is no evidence of how this is enforced or tested.
• Hackathon origin: Suggests a prototype rather than a mature product.

Inference: If the project were to grow, it would need to address ethical concerns around AI-generated personas and data privacy.

1. How many people have used MyTurn beyond yourself?
2. What is the feedback loop like for users after a conversation? Do they return to edit or replay?
3. Are there any plans to monetize or scale this product?
4. How do you ensure that AI-generated characters don’t misrepresent real relationships or individuals?
5. What are your long-term goals for MyTurn beyond the hackathon submission?