AI Engineer Intern (Project: Talking Toy)
Prefer a full account with a password? Sign Up to Apply
Job Description
WHAT YOU WILL BE WORKING ON
Main challenge
You will help build and improve a kid-safe AI chatbot in a children's app that is live in many languages. Your main focus is the LLM-powered chatbot. Its answers must be friendly, consistent, age-appropriate and safe in every supported language. Depending on project needs, you will also work on computer vision (recognizing real-world toys), generative illustration pipelines, and translation and text-to-speech automation.
What it means on a daily basis
• Build and improve AI chatbot experiences: Design, test, and refine prompts and conversation flows to ensure the chatbot delivers natural, consistent, and age-appropriate responses across supported languages.
• Develop AI safety mechanisms: Build and improve safety guardrails, including input/output filters, topic restrictions, and fallback responses to protect children from inappropriate or unsafe content.
• Experiment with and optimize LLMs: Test new models, frameworks, and techniques, evaluate their performance, and optimize API usage for cost, latency, and response quality.
• Build AI automation tools: Develop Python scripts to automate translation and text-to-speech workflows, incorporating validation and quality checks.
• Evaluate, document, and collaborate: Build evaluation datasets, track experiment results, document AI workflows, and work with AI Engineers, Developers, and Content Teams to bring features from prototype to production.
Who you would be working with
• AI engineers, Unity developers, QA and content teams
• Cross-functional and international teams, depending on project needs
Requirements
Department: GAME PROGRAMMING
Function: Information Technology
Experience Level: Internship