Kshitij Grover

Kshitij Grover is the co-founder and chief technology officer of Orb, the billing and revenue-infrastructure company he built with Alvaro Morales and that Adyen acquired for $335 million in 2026. He designs systems that translate unpredictable AI consumption, changing model costs, and automated work into pricing that customers understand and businesses can sustain.

Grover studied computer science and philosophy at Caltech, then worked on infrastructure and led real-time systems at Asana. After participating in the South Park Commons founder fellowship, he founded Orb in 2021 with Morales, another Asana alumnus. His infrastructure background shaped Orb’s approach: retain granular product-usage events and turn them into billing, reporting, prepaid credits, enterprise agreements, and pricing models that can change without compromising financial accuracy.

His case for adaptable billing infrastructure treats monetization as something product teams should iterate on as deliberately as software. That philosophy also extends to security: Grover documented Orb’s SOC 2 Type II compliance program, including access controls and audit logging.

How Grover thinks about AI economics

• Pricing as product design: Pricing determines who can adopt a product, how they use it, and whether they can anticipate its cost. Replit Agent illustrates the tension: charging per checkpoint is legible, while charging for the complexity of completed work may better reflect value but risks surprising customers.

• Outcome-based pricing: Intercom Fin’s support-resolution charges and Chargeflow’s recovered-chargeback fees show how billing can align with recognizable business results. Grover argues that these arrangements work only when outcomes are defined precisely enough to support contracts, customer expectations, and dispute resolution.

• Margin structure over static margins: AI inference costs, model selection, and workload intensity change too quickly for a fixed margin assumption. He emphasizes understanding cost drivers and preventing extreme usage; Cloudflare Workers’ CPU-time billing exemplifies architecture that avoids charging customers for time spent waiting on external models.

• Prepaid credits with spending controls: Credits combine predictable purchasing with flexible consumption, accommodate uneven demand, and simplify discounting across products. Their usability depends on real-time usage visibility, balance alerts, spending limits, and clear protection against unexpected bills.

• Continuous repricing: Changing model economics and product capabilities demand pricing simulations, versioned billing, and careful customer migrations. Repricing also changes sales incentives and account expansion, making monetization an organizational concern as well as an engineering problem.

Grover also argues that specialized enterprise software retains value even as AI makes implementation cheaper: its advantage lies in accumulated domain judgment about edge cases, delayed failures, and workflows spanning multiple teams. His argument for durable enterprise software became particularly relevant when Adyen completed its acquisition of Orb on July 1, 2026, bringing Orb’s billing infrastructure closer to global payments while preserving its standalone product.