Multilingual AI Language-Learning App
Budget / Salary$250–750
TypeFreelance project
LocationRemote
Posted1 hour ago
I’m building an interactive app that lets users practise new languages by holding natural, real-time conversations with an AI tutor. The minimum viable product must handle English, Spanish and Chinese fluently, with follow-on expansion to Korean, Japanese, Thai, Hindi and Portuguese once the core engine is stable.
kind of like the app currently exisitng https://www.speak.com/
Core scope
• Conversation engine that understands and responds contextually in the target language.
• Smooth switch between languages during a session without losing context.
• User interface for text and optional voice chat (mobile or web—whichever stack you are most comfortable with, e.g., React Native, Flutter, Swift/Kotlin).
• Admin dashboard to add new language models as they become available.
• Basic progress tracking: store conversation history, surface common errors, and suggest vocabulary review.
Tech expectations
OpenAI / GPT-style LLM access is fine as the base, but models must be wrapped in a way that lets me retrain or fine-tune per language later. Please keep the architecture modular so additional languages can be slotted in without rewriting the core dialogue manager.
Deliverables
1. Source code in a public Git repository.
2. Working build or hosted demo showing conversations in English, Spanish and Chinese.
3. Brief setup notes and API keys placeholders so I can run everything locally.
Acceptance criteria
• I can start a chat in any of the three launch languages, receive coherent replies, and switch to another language mid-session.
• Latency stays below two seconds for text exchanges on a standard broadband connection.
• All code passes a quick review for readability and clear separation between UI, business logic and AI integration.
If something here needs clarification, flag it early so we can keep the build tight and focused.
kind of like the app currently exisitng https://www.speak.com/
Core scope
• Conversation engine that understands and responds contextually in the target language.
• Smooth switch between languages during a session without losing context.
• User interface for text and optional voice chat (mobile or web—whichever stack you are most comfortable with, e.g., React Native, Flutter, Swift/Kotlin).
• Admin dashboard to add new language models as they become available.
• Basic progress tracking: store conversation history, surface common errors, and suggest vocabulary review.
Tech expectations
OpenAI / GPT-style LLM access is fine as the base, but models must be wrapped in a way that lets me retrain or fine-tune per language later. Please keep the architecture modular so additional languages can be slotted in without rewriting the core dialogue manager.
Deliverables
1. Source code in a public Git repository.
2. Working build or hosted demo showing conversations in English, Spanish and Chinese.
3. Brief setup notes and API keys placeholders so I can run everything locally.
Acceptance criteria
• I can start a chat in any of the three launch languages, receive coherent replies, and switch to another language mid-session.
• Latency stays below two seconds for text exchanges on a standard broadband connection.
• All code passes a quick review for readability and clear separation between UI, business logic and AI integration.
If something here needs clarification, flag it early so we can keep the build tight and focused.
Apply on Freelancer →
Project sourced from Freelancer.com. Applications happen directly on the original platform — we never collect your data.