<aside> <img src="/icons/stars_gray.svg" alt="/icons/stars_gray.svg" width="40px" />
Introducing the AI Transformation Model
How AI native are you? Every company needs AI, but many lack a clear starting point or scaling path. This model shows how to go from manual, fragmented work to AI-powered operations. It mirrors real journeys tracking what changes, expected impact, and the role Notion can play. Every framework has flaws, and no company’s journey is identical to another’s, but we’ve found this model useful for plotting a course.
Created in partnership with Ben Levick (Head of AI & Ops @ Ramp), Geoffrey Litt (Author, MIT researcher, Engineer @ Notion), and many others.
</aside>
Roadmap to becoming AI native
<aside>
</aside>
| Levels | How work changes | How AI operates | Business Impact | Notion's solution |
|---|---|---|---|---|
| Level 1 AI as a Thought Partner | ||||
| Explore ideas, improve decisions. | Individuals use AI to think, write, and analyze faster. Users prompt AI ad-hoc for drafts and suggestions. | • Standalone AI apps | ||
| • General model & memory | ||||
| • User-driven | • Faster output + better decisions (hard to measure consistently) | • AI chatbots | ||
| • Code Q&A | ||||
| • Reply rewriting | ||||
| • Prospects research :notion-logo: | ||||
| • Create & refine docs :notion-logo: | ||||
| • Analyze information :notion-logo: | ||||
| • Translation :notion-logo: | ||||
| Level 2 AI as an Assistant | ||||
| Complete individual tasks faster, save employee time | Individuals use context-aware AI tools for work to complete routine tasks. Users integrate AI into daily work, admins connect data and permissions. | • Task-specific tools | ||
| • Company context access | ||||
| • Embedded + on-demand | ||||
| • Data permissions | • Hours saved per employee per week | |||
| • Quality + velocity gains (ie. less rework, faster onboarding) | ||||
| • Employee satisfaction | • IDE autocomplete + refactor | |||
| • Support assisted replies | ||||
| • Lead scoring/enrichment | ||||
| • AI note-taking :notion-logo: | ||||
| • Search across apps :notion-logo: | ||||
| • Draft documentation :notion-logo: | ||||
| • Handle one-off tasks :notion-logo: | ||||
| • Inbox & calendar triage :notion-logo: | ||||
| Level 3 AI as Teammates | ||||
| Automate repetitive work, increase team efficiency | Teams deploy AI agents that handle recurring workflows. Workflow owners configure agents, with user checkpoints built in. | • Configurable agents | ||
| • Automated runs | ||||
| • Cross-tool execution | ||||
| • Monitoring + versioning | • Team capacity reclaimed (10–40% in recurring work) | |||
| • Reduce process cycles | ||||
| • Scale operations without proportional headcount | • Autonomous coding agents | |||
| • Ticket resolution agents | ||||
| • AI SDRs | ||||
| • Project management agents :notion-logo: | ||||
| • Internal Q&A agents :notion-logo: | ||||
| • Autofill databases :notion-logo: | ||||
| • Connect 3P tools (MCPs) :notion-logo: | ||||
| • Agent analytics + controls :notion-logo: | ||||
| Level 4 AI as the System | ||||
| Run critical workflows, scale organizational capacity | Agents execute high impact, high complexity workflows, iterating continuously. Internal “AI builders” design and manage agents, setting guardrails. | • Multi-agent orchestration | ||
| • Proactive + self-improving | ||||
| • Fully flexible custom tools | ||||
| • Policy, incident/conflict resolution | • Operational leverage (revenue per employee) | |||
| • Scale automation for business-critical process | ||||
| • Faster time-to-market | • Multi-agent for coding | |||
| **** • CX automation | ||||
| • Revenue pipeline orchestration | ||||
| • Agent developer platform :notion-logo: | ||||
| • Agent-powered apps :notion-logo: | ||||
| • Agent orchestration :notion-logo: | ||||
| • Advanced agent governance :notion-logo: |
<aside>
Quick Guide
From manual busywork to automated teamwork
Journey to AI-powered operations
260+ hours back every month, and just getting started
Turning account managers in agent orchestrators