The opportunity
As an engineer at Notion, you’ll help shape core user experiences and accelerate how people discover value in Notion. You'll tackle meaningful challenges with increasing autonomy, crafting code that millions of users will experience.
What you'll do
AI product engineering: building model-powered features end-to-end (UX, APIs, retrieval, orchestration, quality, and reliability)
Model & systems engineering: improving model integration and performance (latency, cost, safety, robustness) and building the infrastructure that supports model serving and experimentation
Evaluation & quality (evals): creating evaluation frameworks and automated/ human-in-the-loop testing to measure and improve model and product quality
Partner with your team to prototype and ship an AI-powered product improvement
Own a scoped productionization project: integrate a new model/technique into an existing workflow, add monitoring + guardrails, or improve latency/cost/reliability.
Contribute to evals and iteration loops: build or extend an evaluation set, run experiments, analyze results, and translate learnings into product or system changes.
What they're looking for
- You have less than two years of engineering experience. You have solid: fundamentals in data structures, algorithms, and distributed systems, with a customer-minded, pragmatic approach to solving problems.
- Expertise building and prototyping: You’re excited to build and iterate quickly, and you’ve started exploring AI/ML through coursework, projects, internships, or hackathons. You’re comfortable learning how different parts of a product fit together (UI, APIs, data), have some familiarity with relational databases like Postgres or MySQL, and can take an idea from prototype to a working feature with guidance.
- Thoughtful problem-solving: You approach problems holistically, starting with a clear and accurate understanding of the context. You think about the implications of what you're building and how it will impact real people's lives. You can navigate ambiguity successfully, decompose complex problems into clean solutions, while also balancing the business impact of what you’re building.
- Impact-driven approach to technology: You see technologies as tools to achieve user impact rather than ends in themselves. You care more about building successful systems that solve real problems than about using specific tech stacks or following trends. You stay current with the latest tools like Cursor, Claude Code, and other AI-assisted development environments, you're pragmatic about choosing the right tool for the job, focusing on what delivers the most value to users and the business.
