SaaStr, Owner.com & Klaviyo reveal their AI agent stacks: 10-20 agents replacing entire teams

The Gist

• SaaStr runs 21+ AI agents with just 2 humans and a dog handling $200M VC fund operations
• Owner.com sees 83% of new customers start journeys via AI products, driving 100% YoY growth
• Klaviyo rebuilt $1.4B revenue product dev process around AI agents from the ground up

• Companies are rebuilding internal operations and customer-facing products around AI agents, replacing human teams.
• AI agents started as simple tools and evolved through continuous commits (600-1,000 commits per agent over months).
• Headless/API-first architecture is critical for AI agent functionality, with Salesforce API being a key example.
• The CEO's hands-on involvement in coding and product development is a common trait among successful AI-driven companies.
• The competitive advantage lies in proprietary feedback loops and training data, not just the AI models themselves.
• Organizational structures are becoming flatter and faster, with teams of agents replacing traditional hierarchies.

• Start small by replacing one boring tool with an AI agent to initiate the transformation.
• Adopt a headless/API-first architecture to enable real-time AI agent functionality.
• Ensure leadership is hands-on in the codebase to drive cultural and technical adoption of AI agents.
• Leverage proprietary data to train AI agents, creating a unique competitive moat.

• 1000 commits (Number of commits required to evolve a tool into an AI agent over a few months.)
• $100M ARR (Owner.com's annual recurring revenue, with 83% of new customers starting their journey via an AI product.)
• 200K+ customers (Klaviyo's customer base, making it a dominant player in B2B e-commerce.)
• 35 founders out of 110 people (Owner.com's team composition in product and engineering, highlighting a flat org structure.)