Founding AI Engineer

Technical Challenges & Role Responsibilities

As part of our team, you'll work on challenges that both move the needle for our mission and align with your strengths—whether that's AI architecture, UI/UX, reinforcement learning, or data engineering.

• Detail

Assessment is our strategic starting point. We need to solve two core technical challenges:

1. Human-AI Alignment: Interpreting RubricsProblem: Rubrics use vague terms like "informed" vs "adequate," causing inconsistent evaluation. Solutions: Rubric unpacking workflows, real-time voice AI training coaches, and reinforcement learning from teacher corrections.

2. Complex Marking: Accurately Evaluating Student WorkProblem: Assessment requires hundreds of decisions across any format (typed, handwritten, graphs) and subject area. Solutions: Multi-agent workflows, OCR/computer vision systems, task-specific analysis modules, and clear feedback conversion.

Both require two-level optimization:

• System level: Breaking problems into flexible multi-agent infrastructures
• Individual agent level: Optimizing each component with evaluation, prompting, fine-tuning, and custom models as needed

Who Should Apply

• You want to be part of a small elite founding team
• You have a huge desire to do something great in the world and can see yourself becoming obsessed with fixing the education system
• You are excited, rather than concerned that you will be living and working out of one shared company house and "working" 24/7 towards one central vision with some of the smartest most driven people in the world
• You want to change the world, and potentially found your own company someday (which we will help you do btw and make the Edexia Mafia >> PayPal Mafia)

Who Should NOT Apply

• You want to have a "work-life balance"
• Note: It's fine if you want to eventually get to a point where you have a work-life balance in 2-5 years. But you at least have to be willing to go through a period where you have no balance at all.
• You don't think you can survive working 24/7 (not necessarily in front of a computer but thinking about ways to solve the problem constantly)
• You are not confident in your ability to keep up
• You largely care about your personal finances
• You just want to solve intellectual puzzles, not real-world problems that will impact the world, sometimes (mostly not) our problems are more boring than intellectual problems and require a level of drive that can’t be built on curiosity alone.

A meeting lands on your calendar

If they’re a yes, I book the chat. You show up — that’s the whole job-hunt.