Lessons from Notion’s Pivot to an AI-First Organization

AI Summary · From a Serial Entrepreneur’s Perspective (The following content is distilled by AI; all opinions belong to the original author; you won’t need to read the full article after this)

This is the AI-era organizational transformation strategy disclosed by Notion’s CEO during a Sequoia interview. Key evidence: retaining 50–60 former founders (A · Tested), with engineering adopting a “barbell model” where senior architects lead 2–3 junior engineers (A · Tested). For those chasing returns, it means: in the AI era, traditional resumes and standardized promotion tracks no longer work; organizations need a mix of highly autonomous entrepreneurs to break through bureaucracy. Action items: hire based on portfolios, not resumes; structure teams as “senior + junior” barbells; and CEOs must personally use AI tools to sense emerging trends.

• Stop asking for resumes during hiring; instead, require candidates to submit actual work samples
• Staff engineering teams along a barbell structure: a 1:2–3 ratio of senior architects to juniors
• Periodically onboard employees with founder backgrounds to prevent bureaucratic drift
• Management must build AI applications themselves; ban reliance on secondhand summaries

1. What Kind of Opportunity Is This

Notion is redefining organizational structure and talent standards for the AI era. The core strategy leverages “Jazz Mode” to dismantle traditional bureaucratic hierarchies and compress middle management through AI. Revenue still flows from SaaS subscriptions, but internally it reduces communication costs and accelerates iteration by restructuring talent—retaining 50–60 former founders and adopting a barbell-shaped engineering team—to maintain a competitive edge as AI features ship.

2. Independent Take

Worth borrowing. The original piece shows that Notion used its barbell model (senior architects mentoring 2–3 newcomers) and portfolio-driven hiring to solve the uncertainty inherent in AI development. My read: this model doesn’t suit every company. It heavily depends on AI agents handling the bulk of repetitive grunt work typically done by junior engineers. For teams without strong AI tooling, tearing out middle management could create execution gaps. The real question is whether AI can genuinely replace “information relay and status synchronization”—and current data suggests it has already partially taken over the coordination role traditionally held by PMs at Notion.

3. Cold-Start Path

First validation move: pick one core feature squad, remove its traditional project-based reporting line, and put one senior architect in direct charge of 2–3 junior engineers, giving them authority to make autonomous decisions using AI agents. Cost: nearly zero—it’s just an internal structural tweak. Timeline: observe for 2–3 months, tracking changes in iteration speed and code quality, with a focus on whether AI tools actually amplify junior output.

4. Biggest Risks and How to Dodge Them

Pitfall one: overestimating AI autonomy while underestimating human collaboration complexity. If junior engineers can’t hit expected output through AI, senior architects get forced back into basic code review, breaking the barbell structure. Mitigation: define clear boundaries for AI-assisted workflows; let seniors handle only architectural judgment and taste (quality control), steering clear of execution details. Pitfall two: founders and executives disconnecting from the front lines. If the CEO only reads secondhand AI summaries instead of building apps themselves, they’ll miss subtle shifts in technology trends and make strategic blunders. Mitigation: force leadership to keep their “builder” identity—require them to personally use the latest AI models to solve one real problem every week.

5. Case Breakdown (What Others Are Doing)

• Demineralizing the talent structure: Notion currently counts 50–60 former founders from other companies among its ranks. These “entrepreneurial DNA” employees are mixed into the organization to combat Big Tech bureaucracy; they kick off new projects, challenge legacy processes, and act as a demineralizing agent for the firm. (My inference: these founders likely joined via acquisition or acqui-hire and operate with high decision-making autonomy.)
• Barbell-model engineering teams: Instead of hiring大量 mid-level engineers, Notion leans into an end-loaded structure. One end: fresh graduates (focused on rapid experimentation, efficient execution, and AI-amplified productivity). The other end: a tiny handful of elite senior architects (driving judgment, taste, and architectural design). The ratio runs about 1 senior architect per 2–3 juniors.
• Reinventing the hiring process: Resumes are dead. The first-round interview skips work history, Big Tech pedigree, and resume length altogether. Instead, candidates must “build something.” The selection rubric has shifted from pure “capability” to “capability × taste × agency,” prioritizing curiosity, optimism, and energy levels.
• Blurring role boundaries: AI product development is treated more like “brewing beer” than “building a bridge”—you can’t precisely control outcomes, so you iterate through constant experimentation. As a result, the lines between PMs, designers, and engineers are fading. Designers, bringing product thinking, user empathy, and visual skills, are starting to absorb traditional PM responsibilities. Marketing has split into Storytelling (closer to product) and Demand Generation (closer to sales), doing away with a single CMO overseeing both.
• Compressing planning cycles: Financial planning still runs quarterly, but product planning now shifts weekly because model capabilities change so fast. CEOs must personally use AI tools (Feel the AI) instead of just watching videos or reading summaries, so they can sense which opportunities are real.

6. Dual-Track Actionability

Cross-border: viable. As a global SaaS benchmark, Notion’s organizational overhaul is built on global collaboration and the widespread adoption of AI tools. Cross-border teams can directly mirror its barbell model and portfolio-driven hiring, using remote collaboration tools to cut management costs. Domestic (China): partially viable. Chinese companies can borrow the “barbell structure” and “portfolio-first” hiring approach, but must account for differences between China’s AI infrastructure and the overseas model ecosystem Notion relies on; local AI tooling will shape how agent workflows get adapted. Additionally, domestic workplace culture is less receptive to flattening hierarchies, so changes need to roll out gradually—bluntly cutting middle management risks leaving a management vacuum.

Original · Investment Internship Studio: Read the full article →