ML/AI Engineer | Notion Global Negotiation Guide

Negotiation DNA: RSU-Heavy (Private, ~$10B Valuation) + Bonus | Workspace AI & LLM Integration | 100M+ Users | NOTION AI PREMIUM

Region	Base Salary	Stock (RSU/4yr)	Bonus	Total Comp
San Francisco	$200K–$260K	$250K–$420K	10–15%	$320K–$460K
New York	$195K–$255K	$245K–$410K	10–15%	$312K–$448K
London	£135K–£175K	£165K–£280K	10–15%	£215K–£305K

Negotiation DNA

Notion ML/AI Engineers build the AI systems powering the connected workspace — from LLM-powered writing assistance to intelligent search to automated database operations for 100M+ users. In February 2026, this is Notion's highest-priority engineering function, as the company integrates AI across every workspace surface to transform how teams create, organize, and discover knowledge. ML/AI Engineers work at the intersection of large language models, retrieval-augmented generation, and product engineering.

As a private company at ~$10B valuation, Notion offers premium equity compensation for AI talent reflecting the strategic importance of the AI roadmap. The competitive landscape for ML/AI engineers is intense in 2026, and Notion competes directly with OpenAI, Anthropic, Google, and other AI-first companies for this talent.

Level Mapping: Notion ML/AI (L3-L5) = Google ML L4-L6 = Meta ML E4-E6 = OpenAI ML = Anthropic ML

🏗️ Notion AI Workspace LLM Integration Lever

Notion's 2026 AI roadmap requires ML/AI Engineers to build production LLM systems that power AI features across every workspace surface — writing assistance with context from the full workspace, intelligent search across documents and databases, and automated workflows triggered by natural language. The ML challenge is unique: models must understand Notion's rich content graph (nested documents, relational databases, wikis) and generate outputs that respect complex permission hierarchies.

ML/AI Engineers must optimize for both quality and latency — users expect sub-second AI responses that maintain workspace context. This requires innovations in RAG architectures, prompt engineering, fine-tuning, and efficient inference for workspace-scale AI features.

Global Levers

1. Workspace-Scale LLM Systems: "I build the LLM-powered features across Notion's workspace — AI writing, search, and automation that understand the full context of documents, databases, and wikis for 100M+ users. Deploying LLMs that understand rich content graphs at this scale is a frontier ML challenge."
2. RAG Architecture Innovation: "Notion's AI requires retrieval-augmented generation across heterogeneous content types — documents, databases, and relational structures. I build RAG systems that maintain quality across these diverse formats while respecting permission boundaries."
3. Sub-Second AI Inference: "Users expect instant AI responses in their workflow. I optimize inference pipelines to deliver high-quality AI outputs in under 500ms while maintaining full workspace context — balancing quality, latency, and cost at scale."
4. OpenAI/Anthropic/Google Competition: "OpenAI, Anthropic, and Google are all competing for ML engineers who can build production AI systems. Notion must offer competitive comp to retain ML talent who can deploy LLMs at workspace scale with product-grade reliability."

Negotiate Up Strategy: "I'm targeting $250K base and $400K RSUs over 4 years with 15% performance bonus for this ML/AI Engineer role. I build the LLM-powered features across Notion's workspace — AI writing, search, and automation for 100M+ users. ML engineers who can deploy production LLM systems at workspace scale are in extreme shortage. I have competing offers from [OpenAI at $450K TC / Google ML at $430K TC / Anthropic at $440K TC]. ML/AI engineers with production LLM and workspace domain expertise are the scarcest talent in 2026." Accept at $225K+ base and $340K+ RSUs.

Evidence & Sources

• [Notion AI — 2026 LLM Integration Roadmap]
• [Notion ML/AI Engineer Comp — Levels.fyi 2025-2026]
• [Notion ~$10B Private Valuation — AI Investment]
• [LLM Engineering — OpenAI/Anthropic/Google Competition 2026]