Software Engineer, Early Career (AI)

About this role

Notion seeks an early-career software engineer to develop AI-powered features using LLMs and embeddings across their collaborative workspace platform. You'll work on end-to-end AI product development, model integration, and evaluation frameworks while taking ownership of projects that impact millions of users.

What you'll do

• Prototype and ship AI-powered product improvements in collaboration with cross-functional teams
• Own productionization projects: integrate new models/techniques, add monitoring and guardrails, optimize latency/cost/reliability
• Build or extend evaluation sets, run experiments, and translate insights into product or system improvements
• Develop model-powered features end-to-end including UX, APIs, retrieval, and orchestration
• Contribute to AI infrastructure and model serving optimization across the platform
• Participate in evaluation and quality assurance for AI features

What they're looking for

• Solid fundamentals in data structures, algorithms, and distributed systems
• Proficiency with React, TypeScript, and Node.js (or willingness to learn)
• Experience with relational databases like Postgres or MySQL
• AI/ML exposure through coursework, projects, internships, or hackathons
• Ability to decompose complex problems and navigate ambiguity
• Proactive communication and high ownership mindset
• Familiarity with AI-assisted development tools (Cursor, Claude Code, etc.)
• Customer-minded and pragmatic problem-solving approach

Benefits

• Work on cutting-edge AI technologies at scale with millions of users
• Mentorship and guidance from experienced AI and product engineers
• Exposure to multiple specialized AI teams (product engineering, model systems, evals)
• Opportunity to shape Notion's AI product vision
• Based in San Francisco with collaborative team environment
• Focus on building technology with real human impact

Likely interview questions

• Tell us about an AI or ML project you've built—what was your role, and what did you learn?
• How would you approach taking an AI prototype from concept to production while managing latency and cost?