🎙️ podcast Analysis February 18, 2026 The a16z Show

Kavak: AI Agent Architecture Replaces 10,000-Employee Workforce Model

AI Infrastructure Automotive Technology Emerging Markets
Conviction HIGH
Risk Profile 1.4/10 (LOW RISK)
Horizon 12-24 months
Signal Snapshot Core Theme: AI Infrastructure

AI tools augment human productivity incrementally

AI agents replace human coordination entirely

Model improvements; Regulatory clarity; Competitive pressure

Executive Summary

Kavak's CEO Carlos García Ottati executed a complete organizational transformation, replacing traditional human-centric operations with AI agents handling 90-95% of customer interactions. The transition required a full year of flat growth (2023) after previously growing 300% annually, but resulted in 4x growth on the other side while reducing headcount from 10,000 to 3,500 employees. This wasn't incremental AI adoption—García Ottati describes deliberately 'burning the ships' and forcing the organization through a painful transition where performance initially deteriorated across all KPIs. The key insight: copilot tools failed because employees wouldn't adopt them, forcing a direct agent-to-customer deployment strategy. Kavak operates across Mexico, Brazil, Chile, Argentina, and the Middle East, managing 10,000 monthly transactions in a market where 40% of used car transactions historically ended in fraud. The company built four separate businesses underneath one consumer experience: e-commerce platform, reconditioning facilities, financing engine, and logistics network. García Ottati's 2017 memo anticipated this transformation, stating they would 'complement our people pipeline with a robot pipeline' years before ChatGPT existed. The operational complexity—managing 60,000 unique SKUs with constant edge cases—made AI deployment both necessary and measurable. This represents a complete reimagining of how vertically integrated businesses can operate in emerging markets where traditional infrastructure doesn't exist.

Key Insights

01 Key Insight
Copilot tools systematically fail in complex operational environments because employees won't adopt them despite having correct information
what Carlos García Ottati said

“We built these copilot tools to give them to our teams so they could just use them and just provide a better customer experience... And we realized very quickly they didn't adopt them.”

Investment Implication Companies building AI tooling should focus on direct agent deployment rather than human-assisted workflows for complex operational tasks

This is a preview. Log in to see the full analysis including investment opportunities, risks, catalysts, and detailed insights.


Next:
The Patronage System: When Fraud Becomes Policy →

Federal prosecutors describe Minnesota's entitlement fraud as 'industrial scale' with $9 billion stolen over seven…

Investment Disclaimer: StackAlpha provides information and analysis tools for educational purposes only. Nothing on this platform constitutes investment advice, and you should not rely solely on this information for investment decisions. Past performance does not guarantee future results. Always consult with qualified financial advisors before making investment decisions. Full Disclaimer