Skip to content
Lenny's 🏦 Vault

Curated AI Knowledge Base Prevents Hallucinations

by Tomer Cohen on December 4, 2025

LinkedIn's Full Stack Builder Model: Rethinking Product Development for the AI Era

LinkedIn's strategic shift to a "Full Stack Builder" model demonstrates how AI can transform traditional product development processes by collapsing organizational complexity and empowering builders to work across the entire stack.

Situation

  • Accelerating pace of change: By 2030, skills required for jobs will change by 70%, and the fastest growing jobs today weren't even on the list a year ago
  • Traditional product development challenges: Over time, LinkedIn's product development process had become increasingly complex with multiple specialized roles, reviews, and handoffs
  • Organizational complexity: The company had evolved into highly specialized functions with micro-specializations within engineering, product, and design
  • Efficiency bottleneck: Simple features required multiple teams, multiple code bases, and multiple sprints just to launch

Actions

Platform Transformation

  • Rearchitected core platforms to make them AI-ready and "reasoned over" by AI tools
  • Created composable UI components with server-side support specifically designed for AI interaction
  • Customized third-party AI tools to work with LinkedIn's unique stack and context

AI Agent Development

  • Built specialized agents for specific functions rather than a single general-purpose agent:
    • Trust Agent: Evaluates potential vulnerabilities and harm vectors in product specs
    • Growth Agent: Analyzes growth opportunities and critiques ideas based on LinkedIn's unique loops and funnels
    • Research Agent: Trained on member personas and historical research to provide contextual insights
    • Analyst Agent: Enables natural language querying of LinkedIn's data graph
    • Maintenance Agent: Automatically fixes failed builds (now handling ~50% of build failures)

Cultural Transformation

  • Established a new "Full Stack Builder" career path that anyone from any function can pursue
  • Replaced the traditional APM program with an "Associate Product Builder" program
  • Created "pods" of cross-functional teams focused on missions rather than functional specialties
  • Updated performance reviews to evaluate AI fluency and cross-functional capabilities
  • Celebrated early wins and showcased examples of people successfully crossing traditional role boundaries

Results

Early Outcomes

  • Teams are saving hours of work per week through AI assistance
  • Higher quality insights and discussions in product development
  • Top performers are embracing the tools most enthusiastically
  • Cross-functional mobility is increasing (e.g., a user researcher successfully transitioning to a growth PM role)
  • Increased experimentation velocity and quality while reducing time to launch

Implementation Challenges

  • Off-the-shelf AI tools never worked without significant customization
  • Initial attempts to give AI access to all internal documents caused hallucinations
  • Different teams gravitated toward different tools, creating standardization challenges
  • Change management proved as important as the technology itself

Key Lessons

  • Curate AI knowledge carefully: "It's not great to just give it access to your drive and say reason all over this knowledge base. It actually does a very poor job understanding importance of the past and putting weights on stuff." Carefully select and clean the knowledge base for AI tools.

  • Invest in platform readiness: Companies must rearchitect their platforms to be AI-ready before seeing significant benefits. This requires substantial upfront investment.

  • Build specialized agents, not general ones: Purpose-built agents for specific functions (trust, growth, research) outperform general-purpose tools, especially when trained on company-specific data.

  • Change management is critical: "It's not enough to give them the tools, you have to build the incentives programs, the motivation, the examples to how you do it." Cultural transformation requires as much attention as technological transformation.

  • Top talent adopts first: The most skilled team members tend to embrace AI tools most quickly, suggesting AI may initially amplify the capabilities of already-strong performers.

  • Align individual and organizational incentives: Frame AI adoption as beneficial both to the organization's competitiveness and to individuals' career development in a rapidly changing job market.

View full episode