Skip to content
Lenny's 🏦 Vault

Performance Reviews Now Assess AI Fluency

by Tomer Cohen on December 4, 2025

The Full Stack Builder Model: Reimagining Product Development for the AI Era

LinkedIn's approach to product development in the age of AI centers on empowering builders to take ideas to market regardless of their traditional role, collapsing organizational complexity and leveraging AI to automate repetitive tasks.

The Case for Change

  • The pace of change now exceeds organizations' ability to respond

    • Skills required for jobs will change by 70% by 2030
    • 70% of today's fastest-growing jobs weren't on the list a year ago
    • Most in-demand jobs are growing by 70%+ year-over-year

  • Traditional product development has become unnecessarily complex:

    • Simple processes expanded into numerous sub-steps (research, reviews, approvals)
    • Process complexity led to organizational complexity
    • Organizational complexity led to micro-specialization
    • Result: Bloated teams, slow iteration, inability to adapt quickly

The Full Stack Builder Model

Core Concept

  • Empower builders to develop experiences end-to-end
  • Combine skills across traditionally distinct domains
  • Create fluid interaction between human and machine
  • Focus on the mission rather than functional specialization

What Humans Should Focus On

  1. Vision: Developing compelling stances about the future
  2. Empathy: Understanding profound unmet needs
  3. Communication: Aligning and rallying others around ideas
  4. Creativity: Generating possibilities beyond the obvious
  5. Judgment: Making high-quality decisions in complex, ambiguous situations

What Should Be Automated

  • Everything else outside those five core traits
  • Research processes
  • Code generation
  • Design implementation
  • Testing and quality assurance
  • Maintenance

Implementation Components

1. Platform Architecture

  • Re-architect core platforms so AI can reason over them
  • Build composable UI components with server-side support
  • Customize third-party AI tools to work with internal systems
  • Significant investment required before benefits accrue

2. AI Tools & Agents

  • Build specialized agents for specific functions:

    • Trust agent: Identifies vulnerabilities and harm vectors
    • Growth agent: Critiques ideas and identifies growth opportunities
    • Research agent: Trained on member personas and past research
    • Analyst agent: Queries the LinkedIn graph without SQL knowledge
    • Maintenance agent: Fixes failed builds (handling ~50% of cases)

  • Key learning: Generic AI tools don't work well with enterprise systems

    • Requires significant customization
    • Need carefully curated "golden examples" rather than access to all data
    • Different teams gravitate to different tools

3. Cultural Transformation

  • Most critical and challenging component
  • Strategies that work:
    • Change performance expectations and evaluations to include AI fluency
    • Pilot success with small "pods" to demonstrate value
    • Celebrate wins and highlight examples in all-hands meetings
    • Create new career paths (Full Stack Builder is now a formal title)
    • Replace APM program with Associate Product Builder program
    • Make tools accessible and encourage feedback

Organizational Structure

  • Move from large functional teams to small, cross-functional "pods"
  • Pods focus on missions/problems rather than functional specialization
  • Teams assemble for specific missions, then reassemble for new challenges
  • Model inspired by Navy SEALs: cross-trained across areas but specialized in missions

Implementation Lessons

  • Top performers adopt new tools first, not those who need the most help

    • They have an innate desire to improve their craft
    • They're willing to invest time before seeing productivity gains

  • Change management is critical

    • Not enough to just provide tools - must build incentives and motivation
    • Show examples of success to create momentum
    • Create permission to experiment without waiting for formal reorganization

  • Patience and investment required

    • Be impatient about the goal but patient about implementation
    • Allocate resources for platform and tool customization
    • Communicate progress continuously, not just the vision

  • Not everyone needs to be a full stack builder

    • Some people prefer specialization
    • System builders who empower full stack builders remain important
    • But fewer specialized roles are needed than in the past
View full episode