jordi romero

All VC-backed tech 2015-2025 generated only $1.44T exit value. Three upcoming IPOs could generate $3T combined. IPOs force conversion to common shares, transparent price discovery, and disclosure of cross-deals (Microsoft-OpenAI-Anthropic loops), cleaning up inflated private valuations with complex structures.

Growth-stage VCs face LP pressure for DPI. General Catalyst's Customer Value Fund (originated at Palantir) finances customer acquisition via non-dilutive, cohort-level revenue share. Provides near-term cash returns to LPs, lower risk than equity growth (no balance sheet lien), and aligns with SaaS unit economics. Now GC's largest fund.

Driven by Palantir's Alex Karp thesis and Anthropic's API trust gap, enterprises are moving from frontier model APIs to specialized models on owned infrastructure (Microsoft AutoGen, local inference) to prevent IP leakage and avoid reinforcement learning on proprietary data — a structural shift accelerated by economic alignment (cost savings + privacy).

Factorial is embedding AI agents directly into its workforce platform to automate complex HR tasks like shift scheduling, performance management, and expenses, using its proprietary compliance and employee data as a competitive moat that prevents customers from exposing sensitive data to external AI tools.

Products evolve through four stages: 1) coding agent adoption, 2) AI-as-service workflows for admins, 3) delegating deterministic logic to goal-seeking agents, 4) full governance delegation where AI writes and maintains policies. Each stage requires stronger data infrastructure and metadata controls.

VCs use complex terms (2x participating preferences, 20% guaranteed IRR, tiered rounds where tier 1 invests at $5M val and brings tier 2 at $20M val same week) to engineer press-release valuations. GP commit % reveals skin in the game. IPOs clean this by forcing common share conversion and full disclosure. Founders often unaware of buried clauses.

Traditional per-seat SaaS pricing breaks when AI agents deliver 10-50x output; model providers charge per token with opaque margins, while application layer must price for indeterminate value creation. Employee-paid tools (Shadow IT) prove product-market fit before enterprise adoption.

Employees pay for AI tools personally (PMF signal) while corporations ban unapproved tools (firing offense at Block). Regulatory schizophrenia: GDPR limits monitoring, NIS2 mandates it. Creates massive opportunity for compliant AI governance platforms that bridge productivity and compliance.