Company Description
NextGenius - a Tech Venture Studio for AI Entrepreneurs, offering project design and development services to startups and innovators. Our key specializations include:
· Artificial Intelligence & Data Science
· Fintech Solutions
· Digital Transformation
· Startup Incubation and Support
Role Description
This is a part-time hybrid role for an AI Tutor Specialist based in Sydney, with flexibility to work from home. Responsibilities include providing educational support in AI-related topics, developing tutoring materials and resources, delivering personalized learning sessions, and tracking student progress to ensure their academic success. Collaboration with the team to improve educational tools and methods is also expected.
Qualifications
Strong knowledge and experience in Artificial Intelligence, Machine Learning, and Data Science
Proven skills in tutoring, mentoring, or teaching; excellent communication and interpersonal skills
Proficiency in programming languages like Python and familiarity with AI/ML frameworks
Ability to adapt teaching methods to diverse learning styles and needs
Organizational skills for managing time efficiently in a hybrid work environment
A proactive approach to continuous learning and staying updated on AI advancements
Reward Hacking Resistance: Develop sophisticated prompt chaining and guardrails to prevent students from jailbreaking the tutor or coercing it into providing direct answers. Implement anti-hallucination and "Socratic method" enforcement layers.
Multi-Platform Integration: Leverage your experience with major LLM providers (OpenAI, Anthropic, Google Gemini, Open-Source models) to select the best model for specific educational tasks (e.g., reasoning vs. summarization) to optimize cost and performance.
Vibe Coding & Prototyping: Utilize agentic coding tools (Claude Code, Cursor, Cline, Codex) to rapidly prototype new interaction models. You will move fast to test whether a new teaching concept works before building it out fully.
AI Agent Development: Build autonomous agents that can evaluate student work, adapt curriculum in real-time, and decide when to escalate a session to a human teacher.
Evaluation & Red Teaming: Create automated evaluation pipelines (RAGAS, LLM-as-a-judge) specifically designed to test for reward hacking scenarios. You will play the role of a "mischievous student" to break your own systems.
Cross-Functional Collaboration: Work closely with curriculum designers to translate K-12 subject matter into structured data that LLMs can reliably utilize.
Bachelor's or advanced degree in Computer Science, Artificial Intelligence, or a related field
Previous experience in educational technology or working with online learning platforms is a plus