About the Role
Research has consistently shown that one-on-one tutoring dramatically outperforms traditional classroom instruction—a finding known as Bloom's 2 Sigma Problem. Despite decades of awareness, no solution has successfully delivered personalized tutoring at scale. Until now.
We are building an AI tutoring system designed to emulate and eventually surpass the best human tutors. You will work on the core AI system that delivers real-time, adaptive learning experiences to learners. If you're excited about building the future of human learning, this role is for you.
Responsibilities
- Build and evolve the core AI tutoring system, including prompt architectures, agentic workflows, and real-time adaptive behaviors.
- Run rigorous experimentation and evaluation, developing custom metrics and frameworks to measure learning effectiveness and improve model reliability.
- Design and implement scalable, maintainable software, integrating AI capabilities with platform systems, content pipelines, and user experience components.
- Collaborate cross-functionally with educators, content creators, and product teams to translate pedagogical goals into technical solutions.
- Prototype and productionize novel model interactions, leveraging emerging research and capabilities to advance the tutor’s intelligence.
- Ensure quality, safety, and observability through monitoring, testing, guardrails, and iterative refinement of model behavior.
- Share knowledge and elevate teammates, contributing to a culture of clarity, simplicity, and continuous learning.
Qualifications
Technical
- 1+ years building complex LLM-based systems, with strong prompt engineering intuition and experience designing agentic workflows.
- Ability to design meaningful evaluation methods, including custom metrics and frameworks when off-the-shelf options fall short.
- Strong software engineering fundamentals, architectural thinking, and fluency in Python; comfortable working with production infrastructure, cloud environments, and CI/CD.
- Actively follows AI research and emerging capabilities, applying new ideas creatively to real problems.
Product & Pedagogy
- Strong empathy for learners and content creators; able to translate pedagogical goals into system behavior and product decisions.
- Ability to reason holistically across AI behavior, content structure, platform features, and user experience.
- Familiarity with learning science or a desire to develop expertise.
Character & Collaboration
- Drawn to difficult, ambiguous problems with no clear blueprint; persistent in exploring and refining solutions.
- Preference for elegance and simplicity in system design.
- Communicates clearly across mediums and enjoys mentoring and collaborating in small, high-ownership teams.